GPCP Version 2.1 Combined Precipitation Data Set Documentation
GPCP Version 2.1 Combined Precipitation Data Set Documentation
George J. Huffman
David T. Bolvin
Laboratory for Atmospheres, NASA Goddard Space Flight Center
and Science Systems and Applications, Inc.
July 6, 2009
What’s New!
July 6, 2009 Version 2.1 has been released, superceding all previous versions, including Version
2. This change was driven by changes to the GPCC gauge analysis, and includes rescaling the
OPI estimates that are used in the pre-SSM/I era. Version 2.1 land estimates are generally higher
than for Version 2, and the pre-SSM/I ocean estimates are also higher. However, the OPI
estimates in the pre-SSM/I era still underestimate the variance that the SSM/I-era estimates
display.
April 9, 2008 The October, November, and December 2007 GPCP data sets have been recomputed
and re-posted. This action became necessary when processing problems in the AIRS
data required a reprocessing for that data set by the AIRS data center; AIRS data are used in the
GPCP products, predominantly at high latitudes. Users are urged to re-acquire these months of
data and discard the previous version.
February 25, 2008 The IDL procedure file read_v2_file.pro is now available to read the header
and entire set of months in a Version 2.1 file into an IDL structure.
June 20, 2006 The multi-satellite precipitation product has been recomputed for the span 1987 2006
to eliminate the inhomogeneity across the 1986/1987 (OPI/SSMI) data boundary over land.
These data are now suitable for long-term studies using the entire data record. These files have
the format gpcp_v2_pms.YYYY, where YYYY is the 4-digit year. This documentation will be
updated shortly to reflect this change to the multi-satellite precipitation product.
June 14, 2006 The GPCP V2 estimates have been recomputed for the span May 2005 -February
2006 due to the recomputation of input AIRS estimates. The AIRS definition of "day" was
inconsistent with the GPCP definition, so the AIRS estimates were recomputed to match the
GPCP. The impact is minimal on the monthly V2 estimates.
April 24, 2006 Beginning with May 2005, AIRS precipitation estimates have replaced the TOVS
estimates at high latitudes because of TOVS instrument termination. The new AIRS data has
been adjusted to match the large-scale bias of the TOVS to maintain homogeneity across the data
boundary.
GPCP V2.1
For simplicity, any distributed dataset that depends on TOVS before May 2005 will utilize
AIRS data in place of the TOVS as of that date. This applies to the datasets ending in "pst",
"est", "ptv", "pms", "ems", "psg", and "esg".
will utilize
AIRS data in place of the TOVS as of that date. This applies to the datasets ending in "pst",
"est", "ptv", "pms", "ems", "psg", and "esg".
Request to Users
The GPCP datasets are developed and maintained with international cooperation and are used by
the worldwide scientific community. To better understand the evolving requirements across the
GPCP user community and to increase the utility of the GPCP product suite, the dataset
producers request that a citation be provided for each publication that uses the GPCP products.
Please email the citation to george.j.huffman@nasa.gov or david.t.bolvin@nasa.gov. Your help
and cooperation will provide valuable information for making future enhancements to the GPCP
product suite.
Contents
1. Data Set Names and General Content
2. Related Projects, Data Networks, and Data Sets
3. Storage and Distribution Media
4. Reading the Data
5. Definitions and Defining Algorithms
6. Temporal and Spatial Coverage and Resolution
7. Production and Updates
8. Sensors
9. Error Detection and Correction
10. Missing Value Estimation and Codes
11. Quality and Confidence Estimates
12. Data Archives
13. Documentation
14. Inventories
15. How to Order and Obtain Information about the Data
Keywords
absolute random error variable
accuracy
AGPI coefficients with missing data
AGPI precipitation product
AIRS
AIRS precipitation product
AIRS quality control
algorithm intercomparison projects
archive and distribution sites
GPCP V2.1
citation list
comparison of Versions 2 and 2.1
contributing centers
data access policy
data file access technique
data set
data set archive
data set creators
data set curator
data set inventory
data set revisions
date
documentation curator
documentation revision history
estimate missing values
GPCP
GPI number of samples product
GPI precipitation product
grid
intercomparison results
IR
IR data correction
known anomalies
known data set issues
known errors
merged SSM/I/TOVS precipitation product
missing months
multi-satellite precipitation product
number of samples variable
obtaining data
OLR
OPI precipitation product
OPI quality control
OPI revisions in 1979 -1981
OPI revision to October 1985
originating machine
pentads
period of record
precipitation variable
production and updates
products
quality index
rain gauge
rain gauge number of samples product
rain gauge precipitation product
rain gauge quality control
comparison of Versions 2 and 2.1
contributing centers
data access policy
data file access technique
data set
data set archive
data set creators
data set curator
data set inventory
data set revisions
date
documentation curator
documentation revision history
estimate missing values
GPCP
GPI number of samples product
GPI precipitation product
grid
intercomparison results
IR
IR data correction
known anomalies
known data set issues
known errors
merged SSM/I/TOVS precipitation product
missing months
multi-satellite precipitation product
number of samples variable
obtaining data
OLR
OPI precipitation product
OPI quality control
OPI revisions in 1979 -1981
OPI revision to October 1985
originating machine
pentads
period of record
precipitation variable
production and updates
products
quality index
rain gauge
rain gauge number of samples product
rain gauge precipitation product
rain gauge quality control
GPCP V2.1
read a month of a product
read a month of byte-swapped product
read the header record
read the monthly climatology
references
satellite-gauge precipitation product
similar data sets
source variable
spatial coverage
spatial resolution
SSM/I
SSM/I composite number of samples product
SSM/I composite precipitation product
SSM/I emission number of samples product
SSM/I emission precipitation product
SSM/I error detection/correction
SSM/I scattering number of samples product
SSM/I scattering precipitation product
standard missing value
technique
temporal resolution
TOVS
TOVS precipitation product
TOVS quality control
units of the variables
read a month of byte-swapped product
read the header record
read the monthly climatology
references
satellite-gauge precipitation product
similar data sets
source variable
spatial coverage
spatial resolution
SSM/I
SSM/I composite number of samples product
SSM/I composite precipitation product
SSM/I emission number of samples product
SSM/I emission precipitation product
SSM/I error detection/correction
SSM/I scattering number of samples product
SSM/I scattering precipitation product
standard missing value
technique
temporal resolution
TOVS
TOVS precipitation product
TOVS quality control
units of the variables
variable
Acronyms
1DD One Degree Daily
AGPI Adjusted GPI
AIP Algorithm Intercomparison Project
AIRS Atmospheric Infrared Sounder
AVHRR Advanced Very High Resolution Radiometer
CPC Climate Prediction Center
CMAP CPC Merged Analysis of Precipitation
DMSP Defense Meteorological Satellite Program
DWD Deutscher Wetterdienst
GARP Global Atmospheric Research Programme
GATE GARP Atlantic Tropical Experiment
Geo Geosynchronous
GEWEX Global Energy and Water Cycle Experiment
GHCN Global Historical Climate Network
GMDC GPCP Merge Development Centre
GMS Geosynchronous Meteorological Satellite
GOES Geosynchronous Operational Environmental Satellites
GPCP V2.1
GPCC Global Precipitation Climatology Centre
GPCP Global Precipitation Climatology Project
GPI Global Precipitation Index
GSFC Goddard Space Flight Center
GSPDC Geostationary Satellite Precipitation Data Centre
HIRS2 High-Resolution Infrared Sounder 2
IR Infrared
lat/lon latitude/longitude
Leo Low-Earth-orbit
MB megabytes
MSU Microwave Sounding Unit
NASA National Aeronautics and Space Administration
NCDC National Climatic Data Center
NCEP National Centers for Environmental Prediction
NESDIS National Environmental Satellite Data and Information Service
NOAA National Oceanic and Atmospheric Administration
OLR Outgoing Longwave Radiation
OPI OLR Precipitation Index
SRDC Surface Reference Data Center
SSM/I Special Sensor Microwave/Imager
Ta Antenna Temperature
Tb Brightness Temperature
TIROS Television Infrared Operational Satellite
TOVS TIROS Operational Vertical Sounder
UTC Universal Coordinated Time (same as GMT, Z)
WCRP World Climate Research Programme
WMO World Meteorological Organization
Global Precipitation Climatology Centre
GPCP Global Precipitation Climatology Project
GPI Global Precipitation Index
GSFC Goddard Space Flight Center
GSPDC Geostationary Satellite Precipitation Data Centre
HIRS2 High-Resolution Infrared Sounder 2
IR Infrared
lat/lon latitude/longitude
Leo Low-Earth-orbit
MB megabytes
MSU Microwave Sounding Unit
NASA National Aeronautics and Space Administration
NCDC National Climatic Data Center
NCEP National Centers for Environmental Prediction
NESDIS National Environmental Satellite Data and Information Service
NOAA National Oceanic and Atmospheric Administration
OLR Outgoing Longwave Radiation
OPI OLR Precipitation Index
SRDC Surface Reference Data Center
SSM/I Special Sensor Microwave/Imager
Ta Antenna Temperature
Tb Brightness Temperature
TIROS Television Infrared Operational Satellite
TOVS TIROS Operational Vertical Sounder
UTC Universal Coordinated Time (same as GMT, Z)
WCRP World Climate Research Programme
WMO World Meteorological Organization
1. Data Set Names and General Content
The *data set* is formally referred to as the "GPCP Version 2.1 Combined Precipitation Data
Set". It is also referred to as the "Version 2.1 Data Set." The Version 2.1 data set supersedes the
previous Version 1, 1c, V2X79, and 2 data sets, which are now considered obsolete.
The current data set provides two final products, the combined satellite-gauge (SG) precipitation
estimate and the combined satellite-gauge precipitation error estimate. The complete data set,
which includes the input and intermediate data files, contains a suite of 27 products providing
monthly, global gridded values of precipitation totals and supporting information for the period
January 1979 – (delayed) present.
Since no single satellite data source spans the entire data record, the product draws upon many
different sources covering different times within the entire data record. The four periods of
differing data coverage are January 1979 -December 1985, January 1986 -June 1987 (and
December 1987), July 1987 -April 2005 (excluding December 1987), and May 2005 -present.
The data contributing to the resulting precipitation estimates for each of these four periods is
GPCP V2.1
discussed in section 5. Substantial attempts have been made to ensure consistency among the
different available input sources.
been made to ensure consistency among the
different available input sources.
The main refereed citation for the data set is Adler et al. (2003; all references are listed in section
13), with the shift from Version 2 to Version 2.1 described in Huffman et al. (2009). The earlier
Version 1 is documented in Huffman et al. (1997), which also appears in Huffman (1997b).
...........................................................................
2. Related Projects, Data Networks, and Data Sets
The *data set creators* are G.J. Huffman, D.T. Bolvin, and R.F. Adler, working in the
Laboratory for Atmospheres, NASA Goddard Space Flight Center, Code 613.1, Greenbelt
Maryland, USA, as the GPCP Merge Development Centre.
...........................................................................
The work is being carried out as part of the Global Precipitation Climatology Project (*GPCP*),
an international project of the WMO/WCRP/GEWEX designed to provide improved long-record
estimates of precipitation over the globe. The GPCP home page is located at
http://www.gewex.org/gpcp.html
...........................................................................
The Version 2.1 Data Set contains data from several *contributing centers*:
1. GPCP Polar Satellite Precipitation Data Centre -Emission (SSM/I emission estimates),
2. GPCP Polar Satellite Precipitation Data Centre -Scattering (SSM/I scattering estimates),
3. GPCP Geostationary Satellite Precipitation Data Centre (GPI and OPI estimates),
4. NASA/GSFC Satellite Research Team (TOVS and AIRS estimates), and
5. GPCP Global Precipitation Climatology Centre (rain gauge analyses),
The final satellite-gauge combination, the single-source input data, and the intermediate satellite-
only combination products are currently being distributed. Some single-source data sets extend
beyond the periods for which they're used in Version 2.1 in their original archival locations.
These input data are only posted by GPCP for months in which they contribute to the final
product.
...........................................................................
The GPCP has sponsored several *algorithm intercomparison projects* (referred to as AIP-1,
AIP-2, and AIP-3) for the purpose of evaluating and intercomparing a variety of satellite
precipitation estimation techniques. As well, the NASA Wetnet Project sponsored several such
projects (referred to as Precipitation Intercomparison Projects, and labeled PIP-1, PIP-2, and
PIP-3). Finally, the WMO/CGMS/IPWG is sponsoring the Project for the Evaluation of High
Resolution Precipitation Products (PEHRPP) which focuses on large-region evaluations over
land at fine scales.
...........................................................................
GPCP V2.1
Only a few *similar data sets* are available. The predecessor monthly GPCP data sets were
produced at GMDC, but are considered superseded by Version 2.1. The Climate Prediction
Center Merged Analysis of Precipitation (CMAP) data set by Xie and Arkin (1996) uses similar
input data and has similar temporal and spatial coverage, but is carried out with a much different
technique. Numerous single-source data sets exist that provide quasi-global coverage; several
are used in this release and are described in Section 5.
...........................................................................
similar data sets* are available. The predecessor monthly GPCP data sets were
produced at GMDC, but are considered superseded by Version 2.1. The Climate Prediction
Center Merged Analysis of Precipitation (CMAP) data set by Xie and Arkin (1996) uses similar
input data and has similar temporal and spatial coverage, but is carried out with a much different
technique. Numerous single-source data sets exist that provide quasi-global coverage; several
are used in this release and are described in Section 5.
...........................................................................
3. Storage and Distribution Media
The current *data set archive* consists of unformatted binary files with ASCII headers. It is
distributed by FTP over the Internet. Each file occupies almost 0.5 MB. The user may also
choose to download the single-source input data and the intermediate combinations.
...........................................................................
4. Reading the Data
The *data file access technique* is the same for all files, regardless of which variable and
estimation technique are related to the file. These files are accessible by standard third-
generation computer languages (FORTRAN, C, etc.).
Each file consists of a 576-byte header record containing ASCII characters (which is the same
size as one row of data), then 12 grids of size 144x72 containing big-endian REAL*4 values.
The header line makes the file nearly self-documenting, in particular spelling out the variable
and technique names, and giving the units of the variable. The header line may be read with
standard text editor tools or dumped under program control. All 12 months of data in the year
are present, even if some have no valid data. Grid boxes without valid data are filled with the
(REAL*4) missing value -99999. The data may be read with standard data-display tools (after
skipping the 576-byte header) or dumped under program control.
...........................................................................
The *originating machine* on which the data files where written is a Silicon Graphics, Inc. Unix
workstation, which uses the "big-endian" IEEE 754-1985 representation of REAL*4 unformatted
binary words. Some CPUs might require a change of representation before using the data.
...........................................................................
It is possible to *read the header record* with most text editor tools, although the size (576
bytes) may be longer than some tools will support. Alternatively, the header record may be
dumped out under program control, as demonstrated in the following programming segment.
The header is written in a KEYWORD=VALUE format, where KEYWORD is a string without
embedded blanks that gives the parameter name, VALUE is a string (potentially) containing
blanks that gives the value of the parameter, and blanks separate each KEYWORD=VALUE
unit. To prevent ambiguity, no spaces or "=" are permitted as characters in PARAMETER, and
“=” is not permitted in VALUE. So, a string followed by “=” signals the start of a new metadata
group.
GPCP V2.1
The sample FORTRAN software to read the header is read_v2.1_header.f, and the sample IDL
procedure is in read_v2.1_file.pro. See ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/software .
............................................................................
, and the sample IDL
procedure is in read_v2.1_file.pro. See ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/software .
............................................................................
It is possible to *read a month of a product*, i.e., one grid of data, with many standard data-
display tools. By design, the 576-byte header is exactly the size of one row of data, so the header
may be bypassed by skipping 576 bytes or 144 REAL*4 data points or one row. Alternatively,
the data may be dumped out under program control as demonstrated in the following
programming segment. Once past the header, there are always 12 grids of size 144x72,
containing big-endian REAL*4 values. All months of data in the year are present, even if some
have no valid data. Grid boxes without valid data are filled with the (REAL*4) "missing" value 99999.
Months in a year that lack data are entirely filled with "missing."
The sample FORTRAN software to read a month of data is read_v2.1_month.f. The sample IDL
procedure to read all months in the year file is in read_v2.1_file.pro. See
ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/software .
...........................................................................
It is also possible to *read a month of byte-swapped product*. The GPCP data are generated
using a Silicon Graphics, Inc. Unix workstation, which uses the "big-endian" IEEE 754-1985
representation of REAL*4 unformatted binary words. To read this data on machines which use
the IEEE "little-endian" format such as Intel-based PCs, the user will need to reverse the order of
the bytes (i.e., byte-swap the data). The code segment below performs this byte swapping. Note
that the code segment below is the same as given above, but with the added feature of swapping
the bytes.
The sample FORTRAN software to read a month of byte-swapped data is
read_v2.1_month_swap.f. The sample IDL procedure to read all months in the year file in
read_v2.1_file.pro automatically handles byte swapping. See
ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/software .
...........................................................................
Standard display tools can be used to *read a monthly climatology* file. No header exists for the
climatology files, so they are each a single big-endian REAL*4 144x72 array. Grid boxes
without valid data are filled with the (REAL*4) "missing" value -99999.
The sample FORTRAN software to read a monthly climatology is read_v2.1_climo.f. See
ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/software .
...........................................................................
5. Definitions and Defining Algorithms
The GPI estimates originally reported on a 2.5°x2.5° lat/lon grid (2.5° GPI) used for the period
January 1986 -December 1996 are provided as accumulations over *pentads*, which are 5-day
periods starting Jan. 1 of each year. That is, pentad 1 covers Jan. 1-5, pentad 2 covers Jan. 6-10,
and pentad 73 covers Dec. 27-31. Leap Day (Feb. 29) is included in pentad 12, which then
GPCP V2.1
covers 6 days. The pentad accumulation period prevents an exact computation of monthly
average for the 2.5° GPI and subsequent products. We assume that a pentad crossing a month
boundary contributes to the statistics in proportion to the fraction of the pentad in the month. For
example, a pentad with 40 images that starts the last day of the month is assumed to contribute 8
images (one-fifth of the full pentad) of rainfall information. The 1°x1° GPI estimates used for
the period January 1997 -present are reported as individual 3-hrly images, and all other input
single-source data fields are provided to GPCP in monthly form.
...........................................................................
exact computation of monthly
average for the 2.5° GPI and subsequent products. We assume that a pentad crossing a month
boundary contributes to the statistics in proportion to the fraction of the pentad in the month. For
example, a pentad with 40 images that starts the last day of the month is assumed to contribute 8
images (one-fifth of the full pentad) of rainfall information. The 1°x1° GPI estimates used for
the period January 1997 -present are reported as individual 3-hrly images, and all other input
single-source data fields are provided to GPCP in monthly form.
...........................................................................
The distributed data set contains 27 *products*, each of which is named by concatenating a
technique name with a variable name. As shown in Table 1, there are 12 precipitation estimation
techniques and four variables, but only 27 of the 35 possible products are considered useful and
archived. Besides product availability, Table 1 displays the abbreviations used for coding the
technique and variable in the file names, the units of the various products, and the currently
distributed products.
NOTE: In general, users wishing to use the "final" combined product should use
the "psg" data files (satellite-gauge combined precipitation product).
Table 1. GPCP Version 2.1 Combined Precipitation Data Set Product List, where * denotes a
distributed product, [ ] gives the abbreviation used for coding the technique or variable in the
file names, and ( ) gives the units of the various products, except Number of Samples, whose
units are displayed in the last column. The technique identifiers for TOVS-related data were not
changed when AIRS data replaced TOVS in those products.
Technique
Variable
Precip
Rate [p]
Absolute
Error [e] Source Number of Samples
(mm/d) (mm/d) [s] [n] (Units)
SSMI Emission [se] * * 55 km images
SSMI Scattering [ss] * * overpass days
SSMI Composite [sc] * * * 55 km images
TOVS (AIRS) [tv] *
SSMI/TOVS(AIRS) Composite [st] * * *
OPI [op] * *
GPI [gp] * * 2.5º images
AGPI [ag] * *
Multi-Satellite [ms] * *
GPCC Gauge [ga] * * * gauges
Satellite-Gauge [sg] * *
For example, the absolute error variable for the multi-satellite technique may be found in files
with "ems" in the name, but there is no product giving the number-of-samples variable for the
multi-satellite technique.
...........................................................................
GPCP V2.1
The *technique* name tells what algorithm was used to generate the product. There are 12 such
techniques in the Version 2.1 Data Set: SSMI Emission, SSMI Scattering, SSMI Composite,
TOVS (AIRS), SSMI/TOVS (AIRS) Composite, OPI, GPI, AGPI, Multi-Satellite,
GHCN+CAMS Rain Gauge, GPCC Rain Gauge, and Satellite-Gauge.
...........................................................................
technique* name tells what algorithm was used to generate the product. There are 12 such
techniques in the Version 2.1 Data Set: SSMI Emission, SSMI Scattering, SSMI Composite,
TOVS (AIRS), SSMI/TOVS (AIRS) Composite, OPI, GPI, AGPI, Multi-Satellite,
GHCN+CAMS Rain Gauge, GPCC Rain Gauge, and Satellite-Gauge.
...........................................................................
The *variable* name tells what parameter is in the product. There are four such variables in the
Version 2.1 Data Set: Precipitation Rate, Absolute Error, Source, and Number of Samples.
...........................................................................
The *precipitation variable* is computed as described under the individual product headings.
All precipitation products have been converted from their original units to mm/d.
..........................................................................
The *SSM/I emission precipitation product* is produced by the Polar Satellite Precipitation Data
Centre -Emission of the GPCP under the direction of L. Chiu, located at the Department of
Geography and GeoInformation Science, George Mason University, Fairfax Virginia, USA, and
the Institute of Space and Earth Information Science, Chinese University of Hong Kong, Hong
Kong, SAR PRC. The Special Sensor Microwave/Imager (SSM/I) data are recorded by selected
Defense Meteorological Satellite Program satellites, and are provided in packed form by Remote
Sensing Systems (Santa Clara, CA). The algorithm applied is the Wilheit et al. (1991) iterative
histogram approach to retrieving precipitation from emission signals in the 19-GHz SSM/I
channel. It assumes a log-normal precipitation histogram and estimates the freezing level from
the 19-and 22-GHz channels. The fit is applied to the full month of data. Individual estimates
on the 2.5ºx2.5º grid occasionally fail to converge. In that case the estimate is set to the simple
average of the 5º precipitation estimates available in the box for the month.
The microwave emission technique infers the quantity of liquid water in a column from the
increased low-frequency observed microwave brightness temperatures. Greater amounts of
liquid water in the column tend to correlate with greater surface precipitation. The algorithm
takes the additional step of fitting a log-normal curve to the month of observations to control
sampling-induced noise. This technique works well over ocean where the surface emissivity is
low and uniform. Over land, however, the emissivity is near one and extremely heterogeneous,
making the scattering algorithm the only choice, so the Wilheit et al. algorithm provides no
estimates.
The available products related to the SSM/I emission precipitation data are provided in Table 1.
...........................................................................
The *SSM/I scattering precipitation product* is produced by the GPCP Polar Satellite
Precipitation Data Centre -Scattering under the direction of R. Ferraro, located in the Center for
Satellite Applications and Research of the NOAA National Environmental Satellite Data and
Information Service (NESDIS/STAR), Washington D.C., USA. The Special Sensor
Microwave/Imager (SSM/I) data are recorded by selected Defense Meteorological Satellite
Program satellites, and are transmitted to NESDIS through the Shared Processing System. The
algorithm applied is based on the Grody (1991) Scattering Index (SI), supplemented by the
GPCP V2.1
Weng and Grody (1994) emission technique in oceanic areas. A similar fall-back approach was
used during the period June 1990 -December 1991 when the 85.5-GHz channels were unusable.
Pixel-by-pixel retrievals are accumulated onto separate daily ascending and descending
0.333ºx0.333º lat/lon grids, then all the grids are accumulated for the month on the 2.5º grid.
-back approach was
used during the period June 1990 -December 1991 when the 85.5-GHz channels were unusable.
Pixel-by-pixel retrievals are accumulated onto separate daily ascending and descending
0.333ºx0.333º lat/lon grids, then all the grids are accumulated for the month on the 2.5º grid.
The microwave scattering technique infers the quantity of hydrometeor ice in a column from the
depressions in the high-frequency 85 GHz channel brightness temperatures. More ice aloft
typically implies more surface precipitation. This relationship is physically less direct than in the
emission technique, but it works equally well over land and ocean whenever deep convection is
important.
The available products related to the SSM/I scattering precipitation data are provided in Table 1.
...........................................................................
The *SSM/I composite precipitation product* is produced as part of the GPCP Version 2.1
Combined Precipitation Data Set by the GPCP Merge Development Centre (see Section 2). The
concept is to take the SSM/I emission estimate over water and the SSM/I scattering estimate over
land. Since the emission technique eliminates land-contaminated pixels individually, a weighted
transition between the two results is computed in the coastal zone. The merger is expressed as
R(emiss) ; N(emiss) ≥
0.75 * N(scat)
R(compos) = N(emiss) * R(emiss) + ( N(scat) -N(emiss) ) * R(scat) (1)
N(scat)
N(emiss) < 0.75 * N(scat)
where R is the precipitation rate; N is the number of samples; composite, emiss, and scat denote
composite, emission, and scattering, respectively; and the 0.75 threshold allows for fluctuations
in the methods of counting samples in the emission and scattering techniques. Note that the
second expression reduces to R(scat) when N(emiss) is zero.
Important Note: The emission and scattering fields used in this merger have been edited to
remove known and suspected artifacts, such as high values in polar regions. These edited fields
may be approximated by using the source variable to mask the emission and scattering fields
contained in this data set. That is, the user may infer that editing must have occurred for points
where the source variable indicates that the scattering or emission (or both) are not used, but the
scattering or emission (or both) values are non-missing.
The available products related to the SSM/I composite precipitation data are provided in Table 1.
............................................................................
The *TOVS precipitation product* is produced by the Satellite Research Team under the
direction of Dr. Joel Susskind, located at NASA Goddard Space Flight Center's Laboratory for
Atmospheres, Greenbelt Maryland, USA. Data from the Television Infrared Operational
Satellite (TIROS) Operational Vertical Sounder (TOVS) instruments aboard the NOAA series of
polar-orbiting platforms are processed to provide a host of meteorological statistics. Susskind
and Pfaendtner (1989) and Susskind et al. (1997) describe the TOVS data processing.
GPCP V2.1
The TOVS precipitation estimates infer precipitation from deep, extensive clouds. The
technique uses a multiple regression relationship between collocated rain gauge measurements
and several TOVS-based parameters that relate to cloud volume: cloud-top pressure, fractional
cloud cover, and relative humidity profile. This relationship is allowed to vary seasonally and
latitudinally. Furthermore, separate relationships are developed for ocean and land.
rom deep, extensive clouds. The
technique uses a multiple regression relationship between collocated rain gauge measurements
and several TOVS-based parameters that relate to cloud volume: cloud-top pressure, fractional
cloud cover, and relative humidity profile. This relationship is allowed to vary seasonally and
latitudinally. Furthermore, separate relationships are developed for ocean and land.
The TOVS data are used for the SSM/I period July 1987 -April 2005 and are provided at the 1º
spatial resolution and at the monthly temporal resolution. The data covering the span July 1987 February
1999 are based on information from two satellites. For the period March 1999 -April
2005, the TOVS estimates are based on information from one satellite due to changes in satellite
data format. In addition, the date span 1-17 February 2004 experienced partial (1st and 17th) or
total (2-16) loss of TOVS data, so AIRS data are used for February 2004.
During their period of use, the TOVS estimates are used for filling in the polar and cold-land
regions in the SSM/I data. The end result is a globally complete "high-quality" precipitation
field for use in adjusting the GPI data.
The available products related to the TOVS precipitation data are provided in Table 1.
............................................................................
The *AIRS precipitation product* is produced by the Satellite Research Team under the
direction of Dr. Joel Susskind, located at NASA Goddard Space Flight Center's Laboratory for
Atmospheres, Greenbelt Maryland, USA. Data from the AIRS instrument aboard the Earth
Observing System Aqua polar-orbiting satellite are processed to provide a host of meteorological
statistics. The processing is similar to the TOVS data processing described earlier.
The AIRS precipitation estimates infer precipitation from deep, extensive clouds. The technique
uses a multiple regression relationship between collocated rain gauge measurements and several
AIRS-based parameters that relate to cloud volume: cloud-top pressure, fractional cloud cover,
and relative humidity profile. This relationship is allowed to vary seasonally and latitudinally.
Furthermore, separate relationships are developed for ocean and land.
The AIRS data are available starting in May 2002, are used for the period May 2005 – present,
and are provided at the 1º spatial resolution and at the monthly temporal resolution. In addition,
the date span 1-17 February 2004 experienced partial (1st and 17th) or total (2-16) loss of TOVS
data, so AIRS data are used for February 2004. The AIRS precipitation estimates have been
bias-adjusted to the TOVS estimates to minimize the TOVS/AIRS data boundary at April/May
2005. Matched histograms of precipitation were computed between the TOVS and AIRS data
for the months January, April, July, and October 2004. These seasonal calibrations are applied
accordingly to the corresponding seasonal months of data after April 2005.
During their period of use, the AIRS estimates are used for filling in the polar and cold-land
regions in the SSM/I data. The end result is a globally complete "high-quality" precipitation
field for use in adjusting the GPI data.
GPCP V2.1
The available products related to the AIRS precipitation data are provided in Table 1.
............................................................................
in Table 1.
............................................................................
The *merged SSM/I/TOVS (AIRS) precipitation product* is produced as part of the GPCP
Version 2.1 Combined Precipitation Data Set by the GPCP Merge Development Centre (see
section 2). In the discussion here, “TOVS” should be understood to include AIRS when those
estimates were used. The coverage of the SSM/I precipitation estimates is limited by the orbit of
the DMSP satellites as well as shortcomings in the microwave technique over cold land. These
holes are filled using the globally complete TOVS data. In the nominal latitude span 40°N-S, the
SSM/I data are used as is. These actual limits on the "as is" band vary over the latitude range
40°-50° North or South depending upon the month of the year. Where there are holes as the
result of cold land, the TOVS data are adjusted to the zonally averaged mean bias of the SSM/I
data and inserted. Just outside of the zone 40ºN-S, the SSM/I and TOVS data are averaged using
equal weighting. Moving further towards the poles where the SSM/I data become less reliable,
the SSM/I-TOVS average is replaced by TOVS data that have been adjusted to a zonally-
averaged presumed bias. In the northern hemisphere, this bias adjustment is anchored on the
equatorward side by the zonal average of the SSM/I-TOVS values anywhere from 50º-60ºN,
depending upon the month of the year. The bias adjustment on the polar side is anchored by the
zonal average of the monthly rain gauge data at 70ºN, with a smooth linear variation in between.
The gauge's zonal average only includes grid boxes for which the gauge "quality index" (defined
in Section 11) is greater than zero. From 70ºN to the North Pole, TOVS data are adjusted to the
bias of the same monthly rain gauge value average at 70ºN. The same procedure is applied in the
southern hemisphere, except the annual climatological rain gauge values are zonally averaged at
70ºS. The monthly values are not used in the Antarctic as the lack of sufficient land coverage
there yields unstable results. Furthermore, the new GPCC analysis lacks data over Antarctica, so
this climatological adjustment is from the previous GPCC Monitoring Product. All seasonal
variations in this description were developed in off-line studies of typical dataset variations, with
the driving criterion being choosing a transition that ensures reasonable performance.
The available products related to the merged SSM/I/TOVS (AIRS) precipitation data are
provided in Table 1.
............................................................................
The *OPI precipitation product* is produced by the Geostationary Satellite Precipitation Data
Centre of the GPCP under the direction of Pingping Xie, located in the Climate Prediction
Center, NOAA National Centers for Environmental Prediction, Washington D.C., USA. The
OPI technique is based on the use of low-Earth orbit satellite outgoing longwave radiation (OLR)
observations (Xie and Arkin 1998). Colder OLR radiances are directly related to higher cloud
tops, which are related to increased precipitation rates. It is necessary to define "cold" locally, so
OLR and precipitation climatologies are computed and a regression relationship is developed for
OLR and precipitation anomalies. In use, the total precipitation inferred is the estimated
anomaly plus the local climatological value. A backup direct OLR-precipitation regression is
used when the anomaly approach yields unphysical values. In this analysis, the precipitation
climatology used to develop the OLR-derived precipitation estimates was based on the GPCP
Version 2.1 satellite-gauge estimates over the time period 1988-2007. The resulting spatially
and temporally varying climatological calibration is then applied to the independent OPI data
GPCP V2.1
covering the span 1979-1987 to fill all months lacking SSM/I data. The OPI data for the first
two satellites (covering January 1979 through August 1981) were given additional adjustments,
described in section 9 under “OPI revisions in 1979-1981” and “OPI revision for October 1985”.
This adjusted OPI data provides a globally complete proxy for the SSM/I data. The available
products related to the OPI precipitation data are provided in Table 1.
............................................................................
1979-1987 to fill all months lacking SSM/I data. The OPI data for the first
two satellites (covering January 1979 through August 1981) were given additional adjustments,
described in section 9 under “OPI revisions in 1979-1981” and “OPI revision for October 1985”.
This adjusted OPI data provides a globally complete proxy for the SSM/I data. The available
products related to the OPI precipitation data are provided in Table 1.
............................................................................
The *GPI precipitation product* is produced by the Geostationary Satellite Precipitation Data
Centre of the GPCP under the direction of Pingping Xie, located in the Climate Prediction
Center, NOAA National Centers for Environmental Prediction, Washington D.C., USA. Each
cooperating geostationary satellite operator (the Geosynchronous Operational Environmental
Satellites, or GOES, United States; the Geosynchronous Meteorological Satellite, or GMS,
Japan, and subsequently the Ministry of Transportation Satellite, MTSat; and the Meteorological
Satellite, or Meteosat, European Community) forward three-hourly "channel 4" ~10.7 micron
thermal infrared (IR) imagery to GSPDC. The global IR rainfall estimates are then generated
from a merger of these data using the GOES Precipitation Index (GPI; Arkin and Meisner,
1987) technique, which relates cold cloud-top area to rain rate.
The GPI technique is based on the use of geostationary satellite IR observations. Colder IR
brightness temperatures are directly related to higher cloud tops, which are loosely related to
increased precipitation rates. From the GATE data, an empirical relationship between brightness
temperature and precipitation rate was developed. For a brightness temperature ≤
235K, a rain
rate of 3 mm/hour is assigned. For a brightness temperature > 235K, a rain rate of 0 mm/hour is
assigned. The GPI works best over space and time averages of at least 250 km and 6 hours,
respectively, in oceanic regions with deep convection.
For the period 1986-March 1998 the GPI data are accumulated on a 2.5ºx2.5º lat/lon grid for
pentads (5-day periods), preventing an exact computation of the monthly average. We assume
that a pentad crossing a month boundary contributes to the statistics in proportion to the fraction
of the pentad in the month. For example, given a pentad that starts the last day of the month, 0.2
(one-fifth) of its samples are assigned to the month in question and 0.8 (four-fifths) of its
samples are assigned to the following month.
Starting with October 1996 the GPI data are accumulated on a 1ºx1º lat/lon grid for individual 3hrly
images. In this case monthly totals are computed as the sum of all available hours in the
month. The Version 2.1 GPI product is based on the 2.5°x2.5° IR data for the period 1988-1996,
and the 1ºx1º beginning in 1997.
In both data sets gaps in geo-IR are filled with low-earth-orbit IR (leo-IR) data from the NOAA
series of polar orbiting meteorological satellites. However, the 2.5ºx2.5º data only contain the
leo-IR used for fill-in, while the 1ºx1º data contain the full leo-IR. The latter allows a more
accurate AGPI (see "AGPI precipitation product"). The Indian Ocean sector routinely lacked
geo-IR coverage until Meteosat-5 was repositioned to that region in June 1998.
See the "IR data correction" and "known data set issues" sections for some additional details on
the GPI data record.
GPCP V2.1
The available products related to the GPI precipitation data are provided in Table 1.
...........................................................................
in Table 1.
...........................................................................
The *AGPI precipitation product* is produced as part of the GPCP Version 2.1 Combined
Precipitation Data Set by the GPCP Merge Development Centre (see section 2). The technique
follows the Adler et al. (1994) Adjusted GPI (AGPI).
During the SSM/I period (starting July 1987), separate monthly averages of approximately
coincident GPI and merged SSM/I/TOVS (AIRS) precipitation estimates are formed by taking
cut-outs of the 3-hourly GPI values that correspond most closely in time to the local overpass
time of the DMSP platform. The ratio of merged SSM/I/TOVS (AIRS) to GPI averages is
computed and controlled to prevent unstable answers. In regions of light precipitation an
additive adjustment is computed as the difference between smoothed merged SSM/I/TOVS
(AIRS) and ratio-adjusted GPI values when the merged SSM/I/TOVS (AIRS) is greater, and
zero otherwise. The spatially varying arrays of adjustment coefficients are then applied to the
full set of GPI estimates. In regions lacking geo-IR data, leo-GPI data are calibrated to the
merged SSM/I/TOVS (AIRS), then these calibrated leo-GPI are calibrated to the geo-AGPI.
This two-step process tries to mimic the information contained in the AGPI, namely the local
bias of the SSM/I and possible diurnal cycle biases in the geo-AGPI. The second step can be
done only in regions with both geo-and leo-IR data, and then smooth-filled across the leo-IR
fill-in. In the case of the 2.5ºx2.5º IR, which lacks leo-IR in geo-IR regions, the missing
calibrated leo-GPI is approximated by smoothed merged SSM/I/TOVS (AIRS) for doing the
calibration to geo-AGPI.
During the pre-SSM/I period (January 1986 -June 1987 and December 1987), the OPI data, as
calibrated by the GPCP satellite-gauge estimates for part of the SSM/I period (1988-2007), are
used as a proxy for the merged SSM/I/TOVS (AIRS) field in the AGPI procedure described for
the SSM/I period. Because the overpass times of the calibrated OPI data are not available, a
controlled ratio between the full monthly calibrated OPI estimates and the full monthly GPI data
is computed. These ratios are then applied to the GPI data to form the AGPI. The additive
constant is computed and applied, when necessary, for light-precipitation regions.
During the pre-SSM/I period January 1979 -December 1985 there is no geo-IR GPI, and
therefore no AGPI. The OPI data, calibrated by the GPCP satellite-gauge estimates for the same
part of the SSM/I period (1988-2007), are used "as is" for the multi-satellite estimates.
The available products related to the AGPI precipitation data are provided in Table 1.
...........................................................................
The *multi-satellite precipitation product* is produced as part of the GPCP Version 2.1
Combined Precipitation Data Set by the GPCP Merge Development Centre (see section 2)
following Huffman et al. (1995). During the SSM/I period, the multi-satellite field as used in the
satellite-gauge combination product (SG) consists of a combination of Geo-AGPI estimates
where available (latitudes 40ºN-S), the weighted combination of the merged SSM/I-TOVS
estimates and the leo-AGPI elsewhere in the 40ºN-S belt, and the merged SSM/I-TOVS data
GPCP V2.1
outside of that zone. Note that “TOVS” in this discussion should be considered to include AIRS
where that data replaced the TOVS estimates. The combination weights are the inverse
(estimated) error variances of the respective estimates. Such weighted combination of SSM/Ithat “TOVS” in this discussion should be considered to include AIRS
where that data replaced the TOVS estimates. The combination weights are the inverse
(estimated) error variances of the respective estimates. Such weighted combination of SSM/ITOVS
and leo-AGPI is done because the leo-IR lacks the sampling to support the full AGPI
adjustment scheme. After use in the SG, the final version of the multi-satellite product is
generated by calibrating the multi-satellite field to the SG, in parallel with the scheme used to
calibrate the OPI to the SG. This step is necessary to ensure consistency between the two
approaches, and was added to GPCP processing 20 June 2006.
During the pre-SSM/I January 1986 -June 1987 and December 1987, the multi-satellite field
consists of a combination of geo-AGPI estimates where available (latitudes 40ºN-S) and the
calibrated OPI estimates elsewhere. The combination weights are the inverse (estimated) error
variances of the respective estimates.
During the pre-SSM/I period January 1979 -December 1985, the OPI data, calibrated by the
GPCP satellite-gauge estimates, are used "as is" for the multi-satellite estimates.
The available products related to the multi-satellite precipitation data are provided in Table 1.
...........................................................................
The *rain gauge precipitation product* for the period 1979 -present is produced by the Global
Precipitation Climatology Centre (GPCC) under the direction of Bruno Rudolf and Udo
Schneider, located in the Deutscher Wetterdienst, Offenbach a.M., Germany (Schneider et al.
2008). Precipitation gauge reports are archived from a time-varying collection of over 70,000
stations around the globe, both from Global Telecommunications System (GTS) reports, and
from other world-wide or national data collections. An extensive quality-control system is run,
featuring an automated screening and then a manual step designed to retain legitimate extreme
events that characterize precipitation. This long-term data collection and preparation activity
now feeds into an analysis that is done in two steps. First, a long-term climatology is assembled
from all available gauge data, focusing on the period 1951-2000. The lack of complete
consistency in period of record for individual stations has been shown to be less important than
the gain in detail, particularly in complex terrain. Then for each month, the individual gauge
reports are converted to deviations from climatology, and are analyzed into gridded values using
a variant of the SPHEREMAP spatial interpolation routine [Willmott et al. 1985]. Finally, the
month’s analysis is produced by superimposing the anomaly analysis on the month’s
climatology.
The GPCC creates multiple products, and two are used in the GPCP Version 2.1. The Full Data
Reanalysis (currently Version 4) is a retrospective analysis that covers the period 1901-2007, and
it is used in GPCP for the span 1979-2007. Thereafter we use the GPCC Monitoring Product
(currently Version 2), which has a similar quality control and the same analysis scheme as the
Full Data Reanalysis, but whose data source is limited to GTS reports. The advantages of these
changes are that 1) we no longer need to use the separate and differently prepared gauge analysis
based on the Global Historical Climate Network and Climate Analysis and Monitoring System
(GHCN+CAMS) for the period 1979-1985, as we did for Version 2, and 2) the numbers of
gauges used are much higher for much of the period of record. When the Full Data Reanalysis is
GPCP V2.1
updated to a longer record we expect to reprocess the GPCP datasets to take advantage of the
improved data. We continue the GPCP’s long-standing practice of correcting all gauge analysis
values for climatological estimates of systematic error due to wind effects, side-wetting,
evaporation, etc., following Legates [1987]. We hope to develop a more modern and detailed
correction for these effects in subsequent versions.
improved data. We continue the GPCP’s long-standing practice of correcting all gauge analysis
values for climatological estimates of systematic error due to wind effects, side-wetting,
evaporation, etc., following Legates [1987]. We hope to develop a more modern and detailed
correction for these effects in subsequent versions.
The available products related to the rain gauge precipitation data are provided in Table 1.
...........................................................................
The *satellite-gauge precipitation product* is produced as part of the GPCP Version 2.1
Combined Precipitation Data Set by the GPCP Merge Development Centre (see section 2) in two
steps (Huffman et al. 1995). Note the subtle point that the multi-satellite (MS) data used here is
the original, in which climatological gauge scaling is approximately included during the pre-
SSM/I era, but not during the SSM/I era.
1a. For each grid box that has less than 65% water coverage on a 5x5-gridbox template:
1b. Average the gauge and MS estimates separately on a 5x5-gridbox template centered on the
box of interest, or a 7x7-gridbox area if there is "too little" data.
1c. Compute the weighted-average gauge to weighted-average MS ratio,
1d. controlling the maximum ratio to be 2 for the weighted-average MS in the range [0,7]
mm/d, 1.25 above 17 mm/d, and linearly tapered in between to suppress artifacts.
1e.
When the ratio exceeds the limit, compute an additive adjustment that is capped at 1.7 mm/d
at zero weighted-average MS and linearly tapers to zero at 7 mm/d. This is intended to
account for the MS badly missing light precipitation.
1f. For all areas with smoothed fractional coverage by water greater than 65%, the ratio is set to
one and the additive adjustment is set to zero.
1g. In each grid box, whether or not there was any adjustment, the gauge-adjusted MS is the
product of the MS and the ratio, added to the additive adjustment.
1h.
In each grid box, whether or not there was any adjustment, the estimated random errors for
both gauge and gauge-adjusted MS are recomputed, using the straight average of the two as
the estimated precipitation value for both calculations. This step prevents inconsistent
results that arise when the random errors are computed with individual precipitation values
that are not close to each other.
2.
In each grid box, whether or not there was any adjustment in step 1, the gauge-adjusted MS
and gauge values are combined in a weighted average, where the weights are the
recomputed inverse (estimated) error variances to form the Satellite-Gauge combination
product.
The available products related to the satellite-gauge precipitation data are provided in Table 1.
...........................................................................
The *absolute random error variable* is produced as part of the GPCP Version 2.1 Combined
Precipitation Data Set by the GPCP Merge Development Centre (see section 2). Following
Huffman (1997a), bias error is neglected compared to random error (both physical and
algorithmic), then simple theoretical and practical considerations lead to the functional form
GPCP V2.1
VAR= H* (rbar+S)*[24 +49* SQRT(rbar)] (2)
Ni
for absolute random error, where VAR is the estimated error variance of an average over a finite
set of observations, H is taken as constant (actually slightly dependent on the shape of the
precipitation rate histogram), rbar is the average precipitation rate in mm/d, S is taken as constant
(approximately SQRT(VAR) for rbar=0), Ni is the number of independent samples in the set of
observations, and the expression in square brackets is a parameterization of the conditional
precipitation rate based on work with the Goddard Scattering Algorithm, Version 2.1 (Adler et
al. 1994) and fitting of (2) to the Surface Reference Data Center analyses (McNab 1995). The
"constants" H and S are set for each of the data sets for which error estimates are required by
comparison of the data set against the SRDC and GPCC analyses and tropical Pacific atoll gauge
data (Morrissey and Green 1991). The computed value of H actually accounts for multiplicative
errors in Ni and the conditional rainrate parameterization (the [ ] term), in addition to H itself.
Table 2 shows the numerical values of H and S. All absolute random error fields have been
converted from their original units of mm/mo to mm/d.
Table 2. Numerical values of H and S constants used to
estimate absolute error for various precipitation estimates.
Technique S (mm/d) H
SSMI Emission [se] 1 3 (55 km images)
SSMI Scattering [ss] 1 3.2 (55 km images)
TOVS (AIRS) [tv] 1 0.0045
OPI [op] 1 0.0045
AGPI [ag] 0.5 0.45 (2.5° images)
Rain Gauge [ga] 0.267 0.0075 (gauges)
For the independent data sets rbar is taken to be the independent estimate of rain itself.
However, when these errors are used in the combination, theory and tests show that the result is a
low bias. Rbar needs to have the same value in all the error estimates; so we estimate it as the
simple average of all rainfall values contributing to the combination. Note that this scheme is
only used in computing errors used in the combination.
The formalism mixes algorithm and sampling error, and should be replaced by a more complete
method when additional information is available from the single-source estimates. However,
when Krajewski et al. (2000) developed and applied a methodology for assessing the expected
random error in a gridded precipitation field, their estimates of expected error agree rather
closely with the errors estimated for the multi-satellite and satellite-gauge combinations.
...........................................................................
The *source variable* is produced as part of the GPCP Version 2.1 Combined Precipitation Data
Set by the GPCP Merge Development Centre (see section 2). It is available for the SSM/I
composite and the SSM/I/TOVS (AIRS) composite techniques and gives the fractional
contribution to the composite by the SSM/I scattering estimate. Referring to (1) in the "SSM/I
composite precipitation product" description, the source SOURCE may be expressed as
GPCP V2.1
0 ;
N(emiss) ≥
0.75 * N(scat)
SOURCE =
( N(scat) -N(emiss) ) N(emiss) < 0.75 * N(scat) (3)
N(scat)
N(SSM/I) + 2 ; SSM/I / TOVS (AIRS) combined
4 ;
TOVS (AIRS)
where N is the number of samples, emiss and scat denote SSM/I emission and scattering,
respectively, N(SSM/I) is the SSM/I source determined from the emission and scattering
components, and the 0.75 threshold allows for fluctuations in the methods of counting samples in
the emission and scattering techniques. Note that the second expression reduces to 1 when
N(emiss) is zero.
...........................................................................
The *number of samples variable* is produced in a variety of units as described under the
individual product headings.
...........................................................................
The *SSM/I emission number of samples product* is provided to the GPCP as the number of
pixels contributing to the grid box average for the month (i.e., the number of "good" pixels). As
part of the Version 2.1 Data Set processing, this number is converted to the number of 55x55 km
boxes that the number of pixels can evenly and completely cover. This conversion provides a
very approximate (over)estimate of the number of independent samples contributing to the
average. The available products related to the SSM/I emission number of samples are provided
in Table 1.
...........................................................................
The *SSM/I scattering number of samples product* is provided to the GPCP as the number of
"overpass days," the count of days in the month that had at least one ascending pass plus days
that had at least one descending pass. As part of the Version 2.1 Data Set processing, this
number is converted to the number of 55x55 km boxes that the number of pixels can evenly and
completely cover. This conversion provides a very approximate (over)estimate of the number of
independent samples contributing to the average. The available products related to the SSM/I
scattering number of samples are provided in Table 1.
...........................................................................
The *SSM/I composite number of samples product* is produced as part of the GPCP Version 2.1
Combined Precipitation Data Set by the GPCP Merge Development Centre (see section 2). Due
to the different units for the SSM/I emission and scattering numbers of samples, it is necessary to
convert at least one before doing the merger. We have chosen to convert overpass days (SSM/I
scattering estimates) to an estimate of complete 55x55 km boxes (our modified units for the
SSM/I emission). In the latitude belt 60°N-S, orbits in the same direction don't overlap on a
single day, and there is an approximate linear relationship between overpass days and 55 km
boxes. Outside that belt the overlaps cause non-linearity, but we ignore it because the general
lack of reliable SSM/I at higher latitudes overwhelms details about the numbers of samples. The
GPCP V2.1
separate numbers of samples for each technique, measured in 55-km boxes, are merged
according to the same formula as the rainfall:
red in 55-km boxes, are merged
according to the same formula as the rainfall:
N(emiss) ; N(emiss) ≥
0.75 * N(scat)
N(compos) = N(emiss) * N(emiss) + ( N(scat) -N(emiss) ) * N(scat) (4)
N(scat)
N(emiss) < 0.75 * N(scat)
where N is the number of samples; composite, emiss, and scat denote composite, emission, and
scattering, respectively; and the 0.75 threshold allows for fluctuations in the methods of counting
samples in the emission and scattering techniques. Note that the second expression reduces to
N(scat) when N(emiss) is zero. The available products related to the SSM/I composite number
of samples are provided in Table 1.
...........................................................................
The *GPI number of samples product* is provided to the GPCP as the number of IR images that
contribute to the 2.5ºx2.5º grid box. For the 2.5ºx2.5º IR data it is provided as the number of
images per pentad (5-day period), while for the 1ºx1º IR data each 3-hrly image is a separate
dataset. For the 2.5ºx2.5º IR data the contribution by pentads that cross month boundaries are
taken to be proportional to the fraction of the pentad in the month.to the fraction of the pentad in
the month. For example, given a pentad that starts the last day of the month, 0.2 (one-fifth) of its
samples are assigned to the month in question and 0.8 (four-fifths) of its samples are assigned to
the following month. The available products related to the GPI number of samples are provided
in Table 1.
..........................................................................
The *rain gauge number of samples product* is provided to the GPCP as the number of stations
providing gauge reports for the month in the 2.5ºx2.5º grid box. The available products related
to the rain gauge number of samples are provided in Table 1.
..........................................................................
The *units of the variables* are given in Table 1 (Section 5) under the entry "Products." In
particular, the precipitation estimates are in mm/day.
..........................................................................
6. Temporal and Spatial Coverage and Resolution
The *date* for a file is the year in which the months it contains occurred. The date for a grid is
the year/month over which the observations were accumulated to form the averages and
estimates. All dates are UTC.
...........................................................................
The *temporal resolution* of the products is one calendar month. The temporal resolution of the
original single-source data sets is also one month, except the GPI data source has pentad (fiveday)
or 3-hrly temporal resolution for the 2.5°x2.5° and 1°x1° IR data sets, respectively. Some
of the single-source data sets are available from other archives at a finer resolution.
GPCP V2.1
...................................................................................................
The *period of record* for the GPCP Version 2.1 Combined Precipitation is January 1979
through the present, delayed a few months for data collection and processing. The start is based
on the availability of the OLR data. The end is based on the availability of input analyses, and is
extended as complete sets of new data arrive. Some of the single-source data sets have longer
periods of record in their original archival sites. The data span for each product available in the
distributed data set is provided in Table 3. Some products are available for longer timespans, but
only the data used in the GPCP Version 2.1 processing is distributed. Data available but not
used in the GPCP Version 2.1 processing are available upon request from the data set creators.
Table 3. GPCP Version 2.1 Combined Precipitation Data Set Product List with data
span coverage in the distributed data set.
Technique/Variable Availability in Distribution
SSMI Emission [se] 07/1987 -11/1987, 01/1988 -present
SSMI Scattering [ss] 07/1987 -11/1987, 01/1988 -present
SSMI Composite [sc] 07/1987 -11/1987, 01/1988 -present
TOVS (AIRS) [tv] 07/1987 -11/1987, 01/1988 – present
TOVS through April 2005, AIRS thereafter
and for 2/2004
SSMI/TOVS(AIRS) Composite [st] 07/1987 -11/1987, 01/1988 -present
OPI [op] 01/1979 -06/1987, 12/1987
GPI [gp] 01/1986 -present
AGPI [ag] 01/1986 -present
Multi-Satellite [ms] 01/1979 -present
GPCC Gauge [ga] 01/1979 -present
Satellite-Gauge [sg] 01/1979 -present
...........................................................................
The *grid* on which each field of values is presented is a 2.5ºx2.5º latitude--longitude
(Cylindrical Equal Distance) global array of points. It is size 144x72, with X (longitude)
incrementing most rapidly West to East from the Prime Meridian, and then Y (latitude)
incrementing North to South. Grid edges are placed at whole-and half-degree values:
First point center = (88.75ºN,1.25ºE)
Second point center = (88.75ºN,3.75ºE)
Last point center = (88.75ºS,1.25ºW)
...........................................................................
The *spatial resolution* of the products is 2.5°x2.5° lat/long, as it was for the original single-
source data sets, except the 1°x1° IR (used starting January 1997). Some of the single-source
data sets are available from other archives at a finer resolution.
...........................................................................
GPCP V2.1
The *spatial coverage* of the products is global in the sense that they are provided on a globaspatial coverage* of the products is global in the sense that they are provided on a global
grid. However, most of the products have meaningful values only on a subset of the grid points.
The single-source products have the largest holes, and the combination products cover
successively more of the globe. See the sensor descriptions (section 8) for additional discussion
of coverage by the single-source products.
...........................................................................
7. Production and Updates
The GPCP is responsible for managing *production and updates* of the GPCP Combined
Precipitation Data Set (WCRP 1986). Version 2.1 is produced by the GPCP Merge
Development Centre (GMDC), located at NASA Goddard Space Flight Center in the Laboratory
for Atmospheres.
Various groups in the international science community are given the tasks of preparing
precipitation estimates from individual data sources, then the GMDC is charged with combining
these into a "best" global product. This activity takes place after real time, at a pace governed
by agreements about forwarding data to the individual centers and activities designed to ensure
the quality in each processing step, and usually happens within three months. The techniques
used to compute the individual and combination estimates are described in section 5.
Updates will be released to (1) extend the data record, (2) take advantage of improved
combination techniques, or (3) correct errors. Updates resulting from the last two cases will be
given new version numbers.
NOTE: The changes described in this section are typical of the changes that are
required to keep the GPCP Combined Precipitation Data Set abreast of current
requirements and science. Users are strongly encouraged to check back
routinely for additional upgrades, and to refer other users to this site rather than
redistributing data that are potentially out of date.
..........................................................................
To date, these *data set revisions* have been implemented from Version 2 to Version 2.1:
1.
The present GPCC analyses replace a combination of the March 1999 version of the
GHCN+CAMS data for the period January 1979 – December 1985, the January 1999 version
of the GPCC Monitoring analysis for 1986-September 1998, and real-time pulls from the
GPCC of Monitoring analyses for subsequent months.
2.
The OPI calibration was changed from using GPCP Version 2 for the period 1988-1995 to
using GPCP Version 2.1 for the period 1988-2007.
3.
The extra adjustment to the first two satellites’ OPI (January 1979 through August 1981) was
recomputed using Version 2.1 and the GPCC Full Data Reanalysis, versus the previous
Version 2 and GHCN+CAMS.
4.
The date span 1-17 February 2004 experienced partial (1st and 17th) or total (2-16) loss of
TOVS data, so AIRS data are used for February 2004 in Version 2.1.
GPCP V2.1
5.
The 5°x5° SSM/I emission-based estimates (i.e., over ocean) for July 1990 – December 1991
were loaded, completing the 5°x5° time series for use as fill-in when the usual 2.5°x2.5°
product failed to converge.
6.
Corrections were made by CPC in the mid-Pacific overlap region between geo-IR satellites
for October and November 1994.
7.
Accumulated minor corrections to the input data sets since the Version 2 computation were
applied in Version 2.1.
..........................................................................
A number of *known data set issues* exist:
1.
The present GPI contains no intersatellite calibration. This is not a serious issue in the AGPI
and combination, although having the intersatellite calibration would provide a better GPI
and at second order refine the AGPI at satellite data boundaries. By contrast, the "official"
NCEP GPI time series has intersatellite calibration for Jan. 1986 -March 1998, then none
thereafter. Tests show that the 40ºN-S oceanic average GPI is about 3% higher for the
intercalibrated data, compared to the non-intercalibrated data.
2.
The present GPI has a 3x3-gridbox smoother applied for non-SSM/I months (Jan. 1986 June
1987, Dec. 1987). Locally, values are different than the non-smoothed version, but
large-area averages should be accurate.
3.
Presently the choice of IR satellite source is strictly by the number of images in the 2.5ºx2.5º
3-hrly pentad IR (used to compute adjustment coefficients), but in the 2.5ºx2.5º pentad IR the
distance to the satellite is also considered (used to compute the AGPI). So, at some locations
nearly equidistant between the two satellites the AGPI is derived for one satellite, but applied
to the other.
NOTE: In the 1ºx1º 3-hrly GPI it is possible for the two satellites to cut in and out on
successive hours. As long as the relative contribution of each is in the same proportion for
both the SSM/I-matched subset and the full data set this is not too important. Using inter-
satellite calibrated data would overcome this issue, although it is likely a second-order effect.
4.
The 1ºx1º IR dataset provides comprehensive leo-IR data while the 2.5ºx2.5º IR only
provides leo-IR in regions lacking geo-IR. The additional data in the 1ºx1º IR allows more
accuracy in estimating the calibration of the SSM/I-calibrated leo-GPI to the geo-AGPI,
causing biases between the 1ºx1º and 2.5ºx2.5º AGPI in leo regions (the Indian Ocean being
the prime case) of up to 15% in the previous Version 1c.
NOTE: Alternatively, a whole different 2.5ºx2.5º pentad low-orbit GPI dataset could be
generated, and then integrated into the system. The improvement over the fix should be only
second-order.
5.
The GMS 2.5ºx2.5º histograms were collected with temperature bin boundaries at half-
degree values, but the 1ºx1º histograms are being collected on whole-degree temperature
boundaries; this causes GPI differences in excess of 10% at 30-40º latitude, and everywhere
the 1ºx1º GPI is smaller. The AGPI largely calibrates out this problem, but if the GPI itself
needs to be consistent, the 235K class could be split in the 1ºx1º histograms in a future
release.
6.
The TOVS precipitation estimates for the SSM/I period July 1987 -February 1999 are based
on two satellites. For February 1999 – April 2005, the TOVS estimates are based on only
one satellite.
GPCP V2.1
7.
TOVS data were partially denied for the period 10-18 September 2001 and cannot be
recovered. As well, various operational issues caused partially or completely missing days of
TOVS data, particularly in the last few months of NOAA-14’s useful life. In a future
reprocessing, partial and completely missing days will be replaced with AIRS data during the
overlap period, May 2002 – April 2005.
8.
The AIRS precipitation estimates are calibrated to approximately match the zonal average
TOVS using the months January, April, July, and October 2004 as the seasonal calibration
months, but regional differences remain.
9.
Beginning with May 2005, AIRS precipitation estimates have replaced the TOVS estimates
at high latitudes because of TOVS instrument termination. The new AIRS data has been
adjusted to match the large-scale bias of the TOVS to maintain homogeneity across the data
boundary. For simplicity, any distributed dataset that depends on TOVS before May 2005
will utilize AIRS data in place of the TOVS as of that date. This applies to the datasets
ending in "pst", "est", "ptv", "pms", "ems", "psg" amd "esg".
10. Every effort has been made to preserve the homogeneity of the Version 2.1 data record.
However, the regional variances inherent in the OPI data are typically smaller than those
encountered in the SSM/I data, so the statistical nature of the Version 2.1 fields will be
different for the pre-SSM/I and SSM/I eras. Future efforts will be directed at minimizing
these differences.
11. The rain gauge data used in the Version 2.1 analysis consists of GPCC Full for the period
1979-2007 and GPCC Monitoring for the period January 2008 -present. Though there is
strong consistency in analysis scheme, quality control, and data sources between the two
analyses, there exists a minimal possibility of a discernible boundary at the cross-over month
for the land precipitation.
12. Every attempt has been made to create an observation-only based precipitation data set.
However, the TOVS estimates rely on numerical model data to initialize the estimation
technique. It is believed that the impact of the numerical model data is minimal on the final
precipitation estimates.
13. Some polar-orbiting satellites can experienced significant drifting of the equator-crossing
time during their period of service. There is no direct effect on the accuracy of the data, but
it is possible that the systematic change in sampling time could introduce biases in the
resulting precipitation estimates. It is unlikely that this issue affects the SSM/I data used for
calibration because the sequence of single satellites used have all stayed within ±1 hour of
the nominal 6 a.m. / 6 p.m. overpass time.
14. The new GPCC climatology/anomaly analysis scheme is intended to perform well where data
are sparse and/or the terrain is complex. Nonetheless, testing remains to show this
everywhere.
...........................................................................
8. Sensors
The Special Sensor Microwave/Imager (*SSM/I*) is a multi-channel passive microwave
radiometer that has flown on selected Defense Meteorological Satellite Program (DMSP)
platforms since mid-1987. The DMSP is placed in a sun-synchronous polar orbit with a period
of about 102 min. The SSM/I provides vertical and horizontal polarization values for 19, 22, 37,
and 85.5 GHz frequencies (except only vertical at 22) with conical scanning. Pixels and scans
GPCP V2.1
are spaced 25 km apart at the suborbital point, except the 85.5-GHz channels are collected at the 85.5-GHz channels are collected at
12.5 km spacing. Every other high-frequency pixel is co-located with the low-frequency pixels,
starting with the first pixel in the scan and the first scan in a pair of scans. The channels have
resolutions that vary from 12.5x15 km for the 85.5 GHz (oval due to the slanted viewing angle)
to 60x75 km for the 19 GHz.
The polar orbit provides nominal coverage over the latitudes 85ºN-S, although limitations in
retrieval techniques prevent useful precipitation estimates in cases of cold land (scattering), land
(emission), or sea ice (both scattering and emission).
The SSM/I is an operational sensor, so the data record suffers the usual gaps in the record due to
processing errors, down time on receivers, etc. Over time the coverage has improved as the
operational system has matured. As well, the first 85.5 GHz sensor to fly degraded quickly due
to inadequate solar shielding. After launch in mid-1987, the 85.5 GHz vertical-and horizontal-
polarization channels became unusable in 1989 and 1990, respectively.
Further details are available in Hollinger et al. (1990).
The SSM/I emission estimates are based on data from the F8 instrument from mid-1987 through
1991, with the F11 being used for 1992 through April 1995, and the F13 thereafter.
...........................................................................
The TIROS Operational Vertical Sounder (*TOVS*) dataset of surface and atmospheric
parameters are derived from analysis of High-Resolution Infrared Sounder 2 (HIRS2) and
Microwave Sounding Unit (MSU) data aboard the NOAA series of polar-orbiting operational
meteorological satellites. The retrieved fields include land and ocean surface skin temperature,
atmospheric temperature and water vapor profiles, total atmospheric ozone burden, cloud-top
pressure and radiatively effective fractional cloud cover, outgoing longwave radiation and
longwave cloud radiative forcing, and precipitation estimate.
For the period January 1979 – March 2005 (used July 1987 – March 2005), the TOVS
precipitation estimates are accumulated on a 1ºx1º lat/lon grid at the monthly temporal
resolution. Due to the estimation technique and the polar orbit of the NOAA satellites, TOVS
provides a globally complete estimate of precipitation. In addition, the date span 1-17 February
2004 experienced partial (1st and 17th) or total (2-16) loss of TOVS data, so AIRS data are used
for February 2004.
For the period January 1979 -February 1999 (used July 1987 – February 1999), the TOVS
estimates are based on two NOAA satellites orbiting in quadrature. Beginning in March 1999,
the TOVS estimates are based on a single NOAA satellite. This occurred as the result of the
failure of NOAA-11.
The various instruments are operational sensors, so the data record suffers the usual gaps in the
record due to processing errors, down time on receivers, sensor failures, etc.
More information can be found in Susskind et al. (1997)
GPCP V2.1
...........................................................................
The Atmospheric Infrared Sounder (*AIRS*) dataset of surface and atmospheric parameters are
derived from analysis of High-Resolution Infrared Sounder data aboard the NASA Aqua polar-
orbiting satellite. The retrieved fields include land and ocean surface skin temperature,
atmospheric temperature and water vapor profiles, total atmospheric ozone burden, cloud-top
pressure and radiatively effective fractional cloud cover, outgoing longwave radiation and
longwave cloud radiative forcing, and precipitation estimate.
For the period April 2005 -present, the AIRS precipitation estimates are accumulated on a 1ºx1º
lat/lon grid at the monthly temporal resolution. In addition, the date span 1-17 February 2004
experienced partial (1st and 17th) or total (2-16) loss of TOVS data, so AIRS data are used for
February 2004. Due to the estimation technique and the polar orbit of the Aqua satellite, AIRS
provides a globally complete estimate of precipitation.
The various instruments are operational sensors, so the data record suffers the usual gaps in the
record due to processing errors, down time on receivers, sensor failures, etc.
...........................................................................
The *OLR* estimates of broadband outgoing longwave radiation are based on an algorithm
applied to the narrow-band IR channels on the Advanced Very High Resolution Radiometer
(AVHRR) aboard the polar-orbiting NOAA series of satellites. Typically two satellites are
available, but occasionally the OLR is based on only one satellite.
The various IR instruments are operational sensors, so the data record suffers the usual gaps in
the record due to processing errors, down time on receivers, sensor failures, etc.
More information can be found in Xie et al. (2000) and Xie and Arkin (1998).
...........................................................................
The infrared (*IR*) data are collected from a variety of sensors. The primary source of IR data
is the international constellation of geosynchronous-orbit meteorological satellites – the
Geosynchronous Operational Environmental Satellites (GOES, United States); the
Geosynchronous Meteorological Satellite (GMS), then the Ministry of Transportation Satellite
(MTSat, both Japanese); and the Meteorological Satellite (Meteosat, European Community).
There are usually two GOES platforms active, GOES-EAST and -WEST, which cover the
eastern and western United States, respectively. Gaps in geosynchronous coverage (most
notably over the Indian Ocean before METEOSAT-5 began imaging there in June 1998) are
filled with IR data from the NOAA-series polar-orbiting meteorological satellites. The
geosynchronous data are collected by scanning (parts of) the earth's disk, while the polar-orbit
data are collected by cross-track scanning. The data are accumulated for processing from full-
resolution (4x8 km) images.
For the period 1986-March 1998 the GPI data are accumulated on a 2.5ºx2.5º lat/lon grid for
pentads (5-day periods). Starting with October 1996 the GPI data are accumulated on a 1ºx1º
lat/lon grid for individual 3-hrly images. In both data sets gaps in geo-IR are filled with low
GPCP V2.1
earth orbit IR (leo-IR) data from the NOAA series of polar orbiting meteorological satellites.
However, the 2.5ºx2.5º data only contain the leo-IR used for fill-in, while the 1ºx1º data contain
the full leo-IR. The GPI product is based on the 2.5ºx2.5º data for the period 1987-1996, and the
1ºx1º beginning in 1997. The boundary is set at January 1997 to avoid placing the boundary
during the 1997-1998 ENSO event.
-IR) data from the NOAA series of polar orbiting meteorological satellites.
However, the 2.5ºx2.5º data only contain the leo-IR used for fill-in, while the 1ºx1º data contain
the full leo-IR. The GPI product is based on the 2.5ºx2.5º data for the period 1987-1996, and the
1ºx1º beginning in 1997. The boundary is set at January 1997 to avoid placing the boundary
during the 1997-1998 ENSO event.
The combination of IR satellites provides near-global coverage, but limitations in retrieval
techniques prevent useful precipitation estimates poleward of about latitude 40°, higher in the
summer hemisphere, and lower in the winter hemisphere.
The various IR instruments are operational sensors, so the data record suffers the usual gaps in
the record due to processing errors, down time on receivers, sensor failures, etc. Most notably,
the GOES series experienced successive failures and replacement over the whole period of
record, and no geo-IR was available in the Indian Ocean sector until METEOSAT-5 was
relocated in that region in mid-1998.
Further details are available in Janowiak and Arkin (1991).
...........................................................................
The *rain gauge* data are quite heterogeneous. Unlike the fairly uniform preparation of satellite
data sets, gauge data sources and qualities are extremely variable. Choice of instrumentation,
including wind-shielding (if any), siting, observing practices, error detection/correction, and
data transmission techniques are all governed by national or regional rules. Typical rain-gauge
models include simple 8-inch cylinders (read manually), weighing (ink trace on graph paper), or
tipping bucket (digital or analog record) devices located in an open area. Reports are generated
manually or automatically and transmitted to a central regional or national site. Most of the rain
gauge reports contributing to the GPCC Monitoring Product were transmitted as SYNOP or
CLIMAT reports on the Global Telecommunications System, amounting to about 7000 reports
per month. In contrast, the Full Data Reanlysis has extensive supplements of national and
regional collections retrieved well after real time, creating a time-varying archive that peaks at
over 40,000 stations in the early 1990’s and declines to less than 10,000 in more recent years.
Further details on the GPCC gauge data are available in Schneider et al. (2008).
...........................................................................
9. Error Detection and Correction
*SSM/I error detection/correction* has several parts. Built-in hot-and cold-load calibration
checks are used to convert counts to Antenna Temperature (Ta). An algorithm has been
developed to convert Ta to Brightness Temperature (Tb) for the various channels (eliminating
cross-channel leakage). As well, systematic navigation corrections are performed. All pixels
with non-physical Tb and local calibration errors are deleted.
Accuracies in the Tb's are within the uncertainties of the precipitation estimation techniques. For
the most part, tests show only small differences among the SSM/I sensors flying on different
platforms.
GPCP V2.1
Some leo-IR/OPI/TOVS satellites experienced significant drifting of the Equator-crossing time
during their period of service. There is no direct effect on the accuracy of the leo-IR/OPI/TOVS
data, but it is possible that the systematic change in sampling time could introduce biases in the
resulting precipitation estimates for regions with strong diurnal variations.
...........................................................................
-IR/OPI/TOVS satellites experienced significant drifting of the Equator-crossing time
during their period of service. There is no direct effect on the accuracy of the leo-IR/OPI/TOVS
data, but it is possible that the systematic change in sampling time could introduce biases in the
resulting precipitation estimates for regions with strong diurnal variations.
...........................................................................
The dominant *IR data correction* is for slanted paths through the atmosphere. Referred to as
"limb darkening correction" in polar-orbit data, or "zenith-angle correction" in geosynchronous-
orbit data (Joyce et al., 2001), this correction accounts for the fact that a slanted path through the
atmosphere increases the chances that (cold) cloud sides will be viewed, rather than (warm)
surface, and raises the altitude dominating the atmospheric emission signal (almost always
lowering the equivalent Tb). In addition, the various sensors have a variety of sensitivities to
the IR spectrum, usually including the 10-11 micron band. Inter-satellite calibration differences
are documented, but they are not implemented in the current version. They are planned for a
future release. The AGPI largely corrects intersatellite calibration, except for small effects at
boundaries between satellites. The satellite operators are responsible for detecting and
eliminating navigation and telemetry errors.
Some IR satellites experienced significant drifting of the equator-crossing time during their
period of service. There is no direct effect on the accuracy of the IR data, but it is possible that
the systematic change in sampling time could introduce biases in the resulting precipitation
estimates for regions with strong diurnal variations.
...........................................................................
The *TOVS quality control* scheme consists of inspection of TOVS precipitation fields for
egregious errors. If errors are detected, the source of the problem is identified and corrected.
Some TOVS satellites experienced significant drifting of the equator-crossing time during their
period of service. There is no direct effect on the accuracy of the TOVS data, but it is possible
that the systematic change in sampling time could introduce biases in the resulting precipitation
estimates for regions with strong diurnal variations.
...........................................................................
The *AIRS quality control* scheme consists of inspection of AIRS precipitation fields for
egregious errors. If errors are detected, the source of the problem is identified and corrected.
...........................................................................
The *OPI quality control* scheme consists of visual inspection of OLR and OLR anomalies for
egregious errors. If errors are detected, the source of the problem is identified and corrected.
...........................................................................
*OPI revisions in 1979-1981* were made to correct apparent calibration-induced biases in the
OPI records from TIROS-N (January 1979 – January 1980) and NOAA-6 (February 1980 –
August 1981). This is true even though the biases in the OLR itself are small (less than 1%), and
this continued to be true for Version 2.1. Accordingly, we re-applied the scheme that was used
GPCP V2.1
in Version 2 to adjust the bias of the first two satellites. The precipitation was averaged for each
satellite separately over all gridboxes having a valid OPI value, at least one gauge/gridbox, and a
gauge estimate of at least 50 mm/month, for all months of TIROS-N (January 1979 – January
1980), NOAA-6 (February 1980 – August 1981), and NOAA-7 (September 1981 – February
1985). The same averaging is applied to the corresponding gauge estimates for the three periods
and compared with the three satellite estimates. The ratios of the averages for each satellite
versus the gauge data were computed. Using the NOAA-7 OPI-gauge ratio as representative,
since it appears to be minimally biased, and assuming that the OPI bias over ocean is similar to
that over land, a ratio correction was applied for all grid boxes to the TIROS-N and NOAA-6
data to match the ratio of the NOAA-7 period. Comparison to an alternative OLR data set [Lee
et al. 2007] shows very similar results, and confirms that biases are consistent between land and
ocean. Nonetheless, the first two satellites still appear to be biased low and will be re-examined
in the next upgrade. The new Version 2.1 adjustments for the TIROS-N and NOAA-6 periods
are +8% and +0.4%, compared to +12% and +3% in Version 2. The GPCP Version 2.1
adjustments are different due to 1) the use of the improved GPCC Full Data Reanalysis
throughout, versus a concatenation of GHCN+CAMS and the prior GPCC Monitoring Products,
and 2) extension of the OLR–GPCP SG calibration period from 1988-1995 in Version 2 to 1988satellite separately over all gridboxes having a valid OPI value, at least one gauge/gridbox, and a
gauge estimate of at least 50 mm/month, for all months of TIROS-N (January 1979 – January
1980), NOAA-6 (February 1980 – August 1981), and NOAA-7 (September 1981 – February
1985). The same averaging is applied to the corresponding gauge estimates for the three periods
and compared with the three satellite estimates. The ratios of the averages for each satellite
versus the gauge data were computed. Using the NOAA-7 OPI-gauge ratio as representative,
since it appears to be minimally biased, and assuming that the OPI bias over ocean is similar to
that over land, a ratio correction was applied for all grid boxes to the TIROS-N and NOAA-6
data to match the ratio of the NOAA-7 period. Comparison to an alternative OLR data set [Lee
et al. 2007] shows very similar results, and confirms that biases are consistent between land and
ocean. Nonetheless, the first two satellites still appear to be biased low and will be re-examined
in the next upgrade. The new Version 2.1 adjustments for the TIROS-N and NOAA-6 periods
are +8% and +0.4%, compared to +12% and +3% in Version 2. The GPCP Version 2.1
adjustments are different due to 1) the use of the improved GPCC Full Data Reanalysis
throughout, versus a concatenation of GHCN+CAMS and the prior GPCC Monitoring Products,
and 2) extension of the OLR–GPCP SG calibration period from 1988-1995 in Version 2 to 19882007
in Version 2.1.
Some OPI satellites experienced significant drifting of the Equator-crossing time during their
period of service. There is no direct effect on the accuracy of the OPI data, but it is possible that
the systematic change in sampling time could introduce biases in the resulting precipitation
estimates for regions with strong diurnal variations.
...........................................................................
*OPI revision to October 1985* was performed to correct for apparent anomalies in the original
OLR data. Unusually cold OLR data produced higher-tna-expected precipitation estimates over
both land and ocean for October 1985. A ratio of GPCP V2.1 SG to OPI was computed using
September and November 1985, adjacent months that exhibited normal behavior, and the
resulting ratio adjustment of 0.919 was applied to the October 1985 OPI data.
...........................................................................
The *rain gauge quality control* scheme for the GPCC gauge data is discussed in Rudolf (1993)
and Section 13. For the most part, quality-control errors are deleted. The largest correctable
error for individual reports is the systematic bias. The use of the Legates (1987) climatological
correction is only an approximate solution, since the correction ought to be applied to the gauges
before averaging. Starting in January 2007 the GPCC began computing an event-by-event
correction new data, which might form a basis for revised corrections in the future. The
availability of rain-gauge reports is extremely variable in space and time, and within a box the
coverage by gauges is often not uniform. As a result, even the "ground truth" of rain gauge data
has non-trivial errors. Analysis values are omitted if the gridbox and all adjacent gridboxes
totally lack gauge sites.
...........................................................................
Seven types of *known errors* are contained in part or all of the current data set, and will be
corrected in a future general re-run. They have been uncovered by visual inspection of the
GPCP V2.1
combined data fields over several years of production, but are considered too minor or
insufficiently understood to provoke an immediate reprocessing. insufficiently understood to provoke an immediate reprocessing.
1.
Limit checks on sea ice contamination in the SSM/I emission estimates continue to be refined
as additional cases are uncovered.
2.
Exact-zero values in marginally snowy land regions (from the SSM/I scattering field) are
probably not reliable, and should simply be "small."
3.
Some leo-IR satellites experience noticeable drift in their equator crossing time, which can
lead to (diurnal) sampling-induced biases of up to 15% in the resulting single-sensor
precipitation estimate.
4.
The AGPI calibration coefficients for the 2.5ºx2.5º IR input (1987-1996) are sometimes
derived on one choice of satellites in regions of overlap between geo satellites, and applied to
another.
5. There is no inter-satellite calibration applied to the GPI.
...........................................................................
Some *known anomalies* in the data set are documented and left intact at the discretion of the
data producers. The current list of anomalies is:
1.
January 2000: The extreme southwestern portion of Greenland the GPCC precipitation
values are unusually high, resulting in correspondingly high values in the combined satellite-
gauge field. According to the GPCC, the high values were the result of near-continuous
precipitation at Nuuk, Greenland (validated by corresponding synoptic reports). The GPCC
believe that the Nuuk gauge precipitation reports are correct in providing greater than normal
precipitation, but perhaps unrealistically so. Eliminating the Nuuk station from the gauge
analysis would produce unrealistically low precipitation values, so it was decided to leave the
station in the analysis. The February 2000 GPCC data shows a similar pattern, but the
precipitation amount at Nuuk is much lower and more in line with surrounding values.
2.
June 1990-December 1991: A fall-back scattering algorithm based on 37 GHz data was used
for the NOAA scattering estimates when both 85.5 GHz channels were inoperable on F08.
The algorithm's sensitivity to precipitation is reduced, particularly light precipitation rates.
3.
August 1993-January 1994: The number of Meteosat-4 IR images decreased to the following
amount:
August 1993: 90% (of nominal number of images)
September 1993: 83%
October 1993: 73%
November 1993: 73%
December 1993: 78%
January 1994: 86%
This drop in the number of IR images caused biased sampling of the coincident SSMI F11
and Meteosat-4 IR observations, resulting in biased monthly microwave-IR calibrations for
the span. Therefore, precipitation in the Meteosat-4 region (~55ºW to ~50ºE longitude) may
be over-or under-estimated depending on location and should be treated as suspect. This is
especically true for the span October-December 1993. As an example, it was discovered that
the biased calibration produced significant overestimation of precipitation in the
GPCP V2.1
Mediterranean Basin. Only the Meteosat-IR sector was affected as the microwave-IR -IR sector was affected as the microwave-IR
calibration is developed and applied locally.
...........................................................................
10. Missing Value Estimation and Codes
There is generally no effort to *estimate missing values* in the single-source data sets, although
a few missing days of gauge data are tolerated in computing monthly values.
...........................................................................
We must compute the *AGPI coefficients with missing data* when leo-GPI data are used to fill
holes in the geo-GPI. In that case, the calibration of the AGPI and SSM/I-calibrated leo-GPI is
computed around the edge of the hole, the calibration coefficients are smoothly filled across the
hole, and applied to the SSM/I-calibrated leo-GPI in the hole. Because the 2.5ºx2.5º IR lacks
leo-GPI in the geo-GPI region, smoothed SSM/I is used to estimate SSM/I-calibrated leo-GPI in
the geo-GPI region. This is not necessary for the 1ºx1º IR because it has leo GPI everywhere.
...........................................................................
All products in the GPCP Version 2.1 Data Set use the *standard missing value* '-99999.' Some
of the single-source data sets possess coded missing values in other archives of the data set.
...........................................................................
Within a GPCP year file, *missing months* are filled entirely with the “standard missing value”
-99999, so that the month number and the position of the month in the file always agree.
...........................................................................
11. Quality and Confidence Estimates
The *accuracy* of the precipitation products can be broken into systematic departures from the
true answer (bias) and random fluctuations about the true answer (sampling), as discussed in
Huffman (1997a). The former are the biggest problem for climatological averages, since they
will not average out. However, on the monthly time scale the low number of samples tends to
present a more serious problem. That is, for most of the data sets the sampling is spotty enough
that the collection of values over one month is not yet representative of the true distribution of
precipitation.
Accordingly, the "random error" is assumed to be dominant, and estimates are computed as
discussed for the "absolute error variable" (section 5). Note that the rain gauge analysis' random
error is just as real as that of the satellite data, even if somewhat smaller. Random error cannot
be corrected.
The "bias error" is not corrected in the SSM/I emission, SSM/I scattering, SSM/I composite, and
GPI precipitation estimates. In the AGPI the GPI is adjusted to the large-scale bias of the SSM/I,
which is assumed lower than the GPI's. As noted in the "satellite-gauge precipitation product"
discussion (section 5), the Multi-Satellite product in the pre-SSM/I (SSM/I) era is (is not)
adjusted to the GPCP climatology (and therefore has gauge influence over land) before use in the
GPCP V2.1
satellite-gauge combination step. However, in both eras the Multi-Satellite product is adjusted
for individual months to the large-scale bias of the Gauge analysis before the combination is
computed. It continues to be the case that biases over ocean cannot be corrected by gauges in the
Multi-Satellite and Satellite-Gauge products. The TOVS, AIRS and OPI data, when used, are
adjusted to the bias of the corresponding SSM/I or rain gauge data, so they are assumed to have
only small bias error.
...........................................................................
-gauge combination step. However, in both eras the Multi-Satellite product is adjusted
for individual months to the large-scale bias of the Gauge analysis before the combination is
computed. It continues to be the case that biases over ocean cannot be corrected by gauges in the
Multi-Satellite and Satellite-Gauge products. The TOVS, AIRS and OPI data, when used, are
adjusted to the bias of the corresponding SSM/I or rain gauge data, so they are assumed to have
only small bias error.
...........................................................................
The single-source estimates have shown reasonable *intercomparison results* in various
intercomparison projects (section 2).
Combinations are difficult to validate as they tend to include data that would otherwise be
independent. An early validation of the old Version 1a data set against the Surface Reference
Data Center analysis yields the statistics in Table 4. Overall, the combination appears to be
working asexpected.
Table 4. Summary statistics for all cells and months comparing the Version 1a
SSM/I composite, Multi-satellite, Gauge, and Satellite-gauge products to the
SRDC analysis for July 1987 – December 1991.
Product Bias (mm/mo) Avg. Diff. (mm/mo) RMS Error (mm/mo)
SSM/I composite 4.03 60.10 88.05
Multi-satellite -5.80 44.20 62.47
Gauge (GPCC) 6.77 18.85 35.11
Satellite-gauge 3.70 20.29 32.98
Krajewski et al. (2000) develop and apply a methodology for assessing the expected random
error in a gridded precipitation field. Their estimates of expected error agree rather closely with
the errors estimated for the multi-satellite and satellite-gauge combinations.
..........................................................................
The *quality index* variable was proposed by Huffman et al. (1997) and developed in Huffman
(1997a) as a way of comparing the errors computed for different techniques. Absolute error
tends to zero as the average precipitation tends to zero, while relative error tends to infinity.
According to (2), the dependence is approximately SQRT(rbar) and 1/SQRT(rbar), respectively.
Thus, it is hard to illustrate overall dependence on sample size with either representation.
However, if one inverts (2) it is possible to get an expression for a number of samples as a
function of precipitation rate and the estimated error variance:
Neg = Hg * (rbarx+Sg)*[1 +10* SQRT(rbarx)] (5)
VARx
where rbarx and VARx are the precipitation rate and estimated error variance for technique X,
Hg and Sg are the values of H and S for the gauge analysis, and Neg is the number of "equivalent
gauges", an estimate of the number of gauges that corresponds to this case. Tests show that Neg
is well-behaved over the range of rbar, largely reflecting the sampling that provided rbarx and
GPCP V2.1
VARx, but also showing differences in the functional form of absolute error over the range of
rbar for different techniques. rbar for different techniques.
Qualitatively, higher Neg denotes more confident answers. Values above 10 are relatively good.
The SSM/I composite estimates tend to have Neg around 1 or 2, while the AGPI has Neg around
3 or 4. The rain gauge analysis runs the whole range from 0 to a few grid boxes in excess of 40.
...........................................................................
An intial *comparison between Versions 2 and 2.1* was developed as part of the transition to
Version 2.1 and has been submitted as part of Huffman et al. (2009). A condensation is provided
here.
The global climatologies for the two versions are quite similar. Differences over oceanic regionsare generally positive and small, representing a compromise between essentially-zero differences
in the SSM/I era and the mean differences during the pre-SSM/I era. In fact, the freckles ofdifference in the oceans mostly correspond to island locations used in the previous GPCCMonitoring Product that have been eliminated from the new Full Data Reanalysis. Land regions
have more substantial differences, mostly due to the mean differences between the versions of
the gauge analyses throughout the period of record. The largest differences, both in magnitudeand extent, occur in northwestern South American and Mesoamerica. The new gauge analysis is
attributing as much as 50% more precipitation to parts of this region, which is characterized by
large gradients in topography. Such gradients typically feature higher precipitation at higherelevations (within the limits of the 2.5° resolution), which the previous GPCC MonitoringProduct tended to miss. Similar tropical topographic regimes are highlighted in Papua NewGuinea, the Himalayas, and along the east coast of the Bay of Bengal. The change in central
Africa is an improvement over the Version 2 data set, in which persistent gaps in gauge coverage
over central Africa coincided with the local maximum in the climatology. Under such
conditions, the previous GPCC analysis scheme, and therefore the GPCP satellite-gauge product,
tended to underestimate the climatological maximum month after month. At higher latitudes themajor increases occur in steep terrain on coasts that intersect storm tracks – the Pacific coasts of
northwestern North America and southern Chile, and New Zealand. Finally, the new GPCC
analysis does not cover Antarctica, so the few gauge sites contained in the previous GPCCanalysis are no longer available. However, the satellite adjustments in the high southern latitudescontinue to be based on the (very approximate) mean gauge precipitation climatology computed
in Version 2 from these gauges as contained in the previous GPCC Monitoring Product.
The time series of global and tropical averages for all, land, and ocean regions give insight intothe aggregate time variation of these differences. Experience has shown that such
regionalization is somewhat sensitive to the choice of regions. Although the most realisticland/water distribution is provided by a threshold for coverage by water of 75%, we wish toprovide a clean “open ocean” comparison of the data sets. Thus, throughout this discussion
“ocean” and “land” regions are defined as having 100% and <100% coverage by water,
respectively. A seven-point boxcar running smoother has been applied to suppress short-interval
noise. As in previous studies, and for both Versions 2 and 2.1, we see that the seasonal cycleover land, primarily driven by the boreal seasons, is almost exactly balanced by changes over theocean on the global scale, with some seasonality apparent in total precipitation for the tropics.
GPCP V2.1
In the pre-SSM/I era (i.e., before mid-1987) the revised scaling for the OPI raises the
oceanic mean for most of the period. The exception occurs in the period January 1979 – August
1981, for the first two satellites. Even though the same bias adjustment procedure is used in bothVersions 2 and 2.1 for these satellites, as described above, we find that the revised OPI scalingand the replacement of the GHCN+CAMS gauge analysis with the new GPCC Full DataReanalysis work together to produce almost no change from Version 2 to Version 2.1, unlike the
later OPI satellites. Over land, Version 2 contains a data boundary at the start of 1986, when the
GHCN+CAMS was replaced by the then-current GPCC Monitoring Product. Reasonable
continuity in the Version 2 time series itself, as well as comparison with the Version 2.1 timeseries, reveals that the change in gauge analysis in January 1998 is relatively unimportant at thetropical or global scale, although locally there can be noticeable differences (not shown).
However, within the GHCN+CAMS record there is an issue. The first few years of GPCPVersion 2 are closer to the corresponding Version 2.1 for the global-average land than any otheryears, confirming earlier impressions that the GHCN+CAMS was making the Version 2 land
estimates in those years somewhat inconsistent with the rest of the record during the use ofGHCN+CAMS (up to 1986).
-SSM/I era (i.e., before mid-1987) the revised scaling for the OPI raises the
oceanic mean for most of the period. The exception occurs in the period January 1979 – August
1981, for the first two satellites. Even though the same bias adjustment procedure is used in bothVersions 2 and 2.1 for these satellites, as described above, we find that the revised OPI scalingand the replacement of the GHCN+CAMS gauge analysis with the new GPCC Full DataReanalysis work together to produce almost no change from Version 2 to Version 2.1, unlike the
later OPI satellites. Over land, Version 2 contains a data boundary at the start of 1986, when the
GHCN+CAMS was replaced by the then-current GPCC Monitoring Product. Reasonable
continuity in the Version 2 time series itself, as well as comparison with the Version 2.1 timeseries, reveals that the change in gauge analysis in January 1998 is relatively unimportant at thetropical or global scale, although locally there can be noticeable differences (not shown).
However, within the GHCN+CAMS record there is an issue. The first few years of GPCPVersion 2 are closer to the corresponding Version 2.1 for the global-average land than any otheryears, confirming earlier impressions that the GHCN+CAMS was making the Version 2 land
estimates in those years somewhat inconsistent with the rest of the record during the use ofGHCN+CAMS (up to 1986).
The total, land, and ocean averages for each of the Versions are given in Table 5. The
global and tropical regions are consistent in showing modestly higher values in Version 2.1, with
essentially all of the change occurring over land (and coast, since the threshold is 100%).
Table 5 Global-and tropical-average land, ocean, and total precipitation for Versions
2.1 and 2 in mm/d. The percentage increase of Version 2.1 over Version 2 is given in
parentheses. “Ocean” and “land” regions are defined by 100% and <100% coverage
by water.
1979-2007 Global
90°N-90°S
Tropical25°N-25°S
Version 2 2.1 2 2.1
Land and Ocean 2.62 2.68 (+2%) 3.12 3.22 (+3%)
Land 2.39 2.53 (+6%) 3.49 3.73 (+7%)
Ocean 2.78 2.78 (+0%) 2.88 2.88 (+0%)
One convenient way to summarize time changes in the data sets is to compute the long-
term linear rate of change for each grid box. Note that we compute the linear change statisticwith no assumption or implication of a particular dynamic or secular trend. Furthermore, thechange in input satellite data from OPI to SSM/I led us to compute the linear changes both forthe entire data set, and for the SSM/I era (1988-2007). The global-and tropical-average linear
changes are listed in Table 6. In general, there is consistency both between the longer andshorter period results and between the Version 2 and 2.1 results. The more precise, somewhat
shorter, and more recent SSM/I-era data mostly show larger trend values than the entire data
record, while Version 2.1 shows trends closer to zero than Version 2. The increase in linear
change from Version 2 to Version 2.1 for global ocean across the entire data set, which is the
only such increase in Table 6, is driven by the somewhat questionable behavior of the OPI in thefirst 2.5 years of the data record. Large revisions to the linear change over land from Version 2
to 2.1 tend to be focused in the regions previously noted for having large mean differences
GPCP V2.1
between the two Versions. As stated previously, the statistics under discussion are somewhatsensitive to the definition of land and ocean.
. As stated previously, the statistics under discussion are somewhatsensitive to the definition of land and ocean.
Table 6 Global-and tropical-average linear changes in mm/d/decade a) for the entire
study period (1979-2007), and b) for the SSM/I era (1988-2007) . “Ocean” and
“land” regions are defined by 100% and <100% coverage by water.
a)
1979-2007 Global
90°N-90°S
Tropical25°N-25°S
Version 2 2.1 2 2.1
Land and Ocean +0.0115 +0.0069 +0.0480 +0.0252
Land +0.0018 –0.0018 +0.0432 -0.0030
Ocean +0.0184 +0.0199 +0.0511 +0.0438
b)
1988-2007 Global
90°N-90°S
Tropical25°N-25°S
Version 2 2.1 2 2.1
Land and Ocean +0.0343 +0.0169 +0.0714 +0.0497
Land +0.0630 +0.0252 +0.1327 +0.0889
Ocean +0.0144 +0.0111 +0.0308 +0.0238
..........................................................................
12. Data Archives
The *archive and distribution sites* for the GPCP Version 2.1 Combined Precipitation Data Set
are as follows:
Mr. David Smith
World Data Center A (WDC-A)
National Climatic Data Center (NCDC)
Rm 120
151 Patton Ave.
Asheville, NC 28801-5001 USA
Phone: 828-271-4053
Fax: 828-271-4328
Internet: david.p.smith@noaa.gov
WDC-A Home Page: http://lwf.ncdc.noaa.gov/oa/wmo/wdcamet-ncdc.html
Dr. George J. Huffman
Code 613.1
NASA Goddard Space Flight Center
Greenbelt, MD 20771 USA
Phone: 301-614-6308
GPCP V2.1
Fax: 301-614-5492
Internet: george.j.huffman@nasa.gov
MAPB Precipitation Page: http://precip.gsfc.nasa.gov
301-614-5492
Internet: george.j.huffman@nasa.gov
MAPB Precipitation Page: http://precip.gsfc.nasa.gov
Independent archive and distribution sites exist for the single-source data sets, and a current list
may be obtained by contacting Mr. Smith at NCDC.
..........................................................................
13. DOCUMENTATION
The *documentation curator* is:
David T. Bolvin
Code 613.1
NASA Goddard Space Flight Center
Greenbelt, MD 20771 USA
Phone: 301-614-6323
Fax: 301-614-5492
Internet: david.t.bolvin@nasa.gov
MAPB Precipitation Page: http://precip.gsfc.nasa.gov
..........................................................................
The *documentation revision history* is:
December 2, 1999 Draft 1 by GJH
January 23, 2000 Final by DTB
March 10, 2000 Rev.1 by DTB
April 28, 2000 Rev.2 by GJH
May 22, 2000 Rev.3 by DTB
August 8, 2000 Rev.4 by DTB
August 24, 2000 Rev.5 by DTB
October 5, 2000 Rev.6 by DTB
December 7, 2000 Rev.7 by DTB
February 8, 2001 Rev.8 by DTB
February 23, 2001 Rev.9 by DTB
March 14, 2001 Rev.10 by DTB
March 28, 2001 Rev.11 by DTB
April 18, 2001 Rev.12 by DTB
May 31, 2001 Rev.13 by DTB
June 4, 2001 Rev.14 by DTB
June 28, 2001 Rev.15 by DTB
August 1, 2001 Rev.16 by DTB
August 18, 2001 Rev.17 by DTB
September 4, 2001 Rev.18 by DTB
October 17, 2001 Rev.19 by DTB
November 2, 2001 Rev.20 by DTB
GPCP V2.1
December 14, 2001 Rev.21 by DTB
February 5, 2002 Rev.22 by DTB
March 29, 2002 Rev.23 by DTB
April 4, 2002 Rev.24 by DTB
May 22, 2002 Rev.25 by DTB
July 31, 2002 Rev.26 by DTB
August 22, 2002 Rev.27 by DTB
August 30, 2002 Rev.28 by DTB
September 10, 2002 Rev.29 by DTB
September 11, 2002 Rev.30 by GJH
September 26, 2002 Rev.31 by DTB
October 31, 2002 Rev.32 by DTB
January 10, 2003 Rev.33 by DTB
January 29,2003 Rev.34 by DTB
March 14, 2003 Rev.35 by DTB
July 28, 2003 Rev.35 by DTB
January 14, 2004 Rev.36 by DTB
January 27, 2004 Rev.37 by DTB
February 3, 2004 Rev.38 by DTB
February 19, 2004 Rev.39 by DTB
February 25, 2004 Rev.40 by DTB
March 15, 2004 Rev.41 by DTB
April 21, 2004 Rev.42 by DTB
July 14, 2004 Rev.43 by DTB
July 16, 2004 Rev.44 by DTB
August 2, 2004 Rev.45 by DTB
October 27, 2004 Rev.46 by DTB
November 3, 2004 Rev.47 by DTB
November 19, 2004 Rev.48 by DTB
December 13, 2004 Rev.49 by DTB
January 11, 2005 Rev.50 by DTB
January 21, 2005 Rev.51 by DTB
February 18, 2005 Rev.52 by DTB
March 22, 2005 Rev.53 by DTB
April 24, 2005 Rev.54 by DTB
June 6, 2005 Rev.55 by DTB
June 22, 2005 Rev.56 by DTB
November 29, 2005 Rev.57 by DTB
April 14, 2006 Rev.58 by DTB
April 24, 2006 Rev.59 by DTB
June 1, 2006 Rev.60 by GJH/DTB
June 14, 2006 Rev.61 by DTB
June 20, 2006 Rev.62 by DTB
June 28, 2006 Rev.63 by DTB
August 14, 2006 Rev.64 by DTB
September 9, 2006 Rev.65 by DTB
Rev.21 by DTB
February 5, 2002 Rev.22 by DTB
March 29, 2002 Rev.23 by DTB
April 4, 2002 Rev.24 by DTB
May 22, 2002 Rev.25 by DTB
July 31, 2002 Rev.26 by DTB
August 22, 2002 Rev.27 by DTB
August 30, 2002 Rev.28 by DTB
September 10, 2002 Rev.29 by DTB
September 11, 2002 Rev.30 by GJH
September 26, 2002 Rev.31 by DTB
October 31, 2002 Rev.32 by DTB
January 10, 2003 Rev.33 by DTB
January 29,2003 Rev.34 by DTB
March 14, 2003 Rev.35 by DTB
July 28, 2003 Rev.35 by DTB
January 14, 2004 Rev.36 by DTB
January 27, 2004 Rev.37 by DTB
February 3, 2004 Rev.38 by DTB
February 19, 2004 Rev.39 by DTB
February 25, 2004 Rev.40 by DTB
March 15, 2004 Rev.41 by DTB
April 21, 2004 Rev.42 by DTB
July 14, 2004 Rev.43 by DTB
July 16, 2004 Rev.44 by DTB
August 2, 2004 Rev.45 by DTB
October 27, 2004 Rev.46 by DTB
November 3, 2004 Rev.47 by DTB
November 19, 2004 Rev.48 by DTB
December 13, 2004 Rev.49 by DTB
January 11, 2005 Rev.50 by DTB
January 21, 2005 Rev.51 by DTB
February 18, 2005 Rev.52 by DTB
March 22, 2005 Rev.53 by DTB
April 24, 2005 Rev.54 by DTB
June 6, 2005 Rev.55 by DTB
June 22, 2005 Rev.56 by DTB
November 29, 2005 Rev.57 by DTB
April 14, 2006 Rev.58 by DTB
April 24, 2006 Rev.59 by DTB
June 1, 2006 Rev.60 by GJH/DTB
June 14, 2006 Rev.61 by DTB
June 20, 2006 Rev.62 by DTB
June 28, 2006 Rev.63 by DTB
August 14, 2006 Rev.64 by DTB
September 9, 2006 Rev.65 by DTB
GPCP V2.1
September 26, 2006 Rev.66 by DTB
October 26, 2006 Rev.67 by DTB
December 14, 2006 Rev.68 by DTB
January 10, 2007 Rev.69 by DTB
January 22, 2007 Rev.70 by DTB
February 16, 2007 Rev.71 by DTB
March 20, 2007 Rev.72 by DTB
April 24, 2007 Rev.73 by DTB
May 16, 2007 Rev.74 by DTB
June 15, 2007 Rev.75 by DTB
June 28, 2007 Rev.76 by DTB
August 6, 2007 Rev.77 by DTB
August 24, 2007 Rev.78 by DTB
October 10, 2007 Rev.79 by DTB
October 26, 2007 Rev.80 by DTB
October 31, 2007 Rev.81 by GJH/DTB convert to MSWord-based PDF
February 26, 2008 Rev.82 by GJH
April 9, 2008 Rev.83 by GJH
December 8, 2008 Rev.84 by GJH
July 6, 2009 Rec.85 by GJH/DTB Version 2.1
..........................................................................
Rev.66 by DTB
October 26, 2006 Rev.67 by DTB
December 14, 2006 Rev.68 by DTB
January 10, 2007 Rev.69 by DTB
January 22, 2007 Rev.70 by DTB
February 16, 2007 Rev.71 by DTB
March 20, 2007 Rev.72 by DTB
April 24, 2007 Rev.73 by DTB
May 16, 2007 Rev.74 by DTB
June 15, 2007 Rev.75 by DTB
June 28, 2007 Rev.76 by DTB
August 6, 2007 Rev.77 by DTB
August 24, 2007 Rev.78 by DTB
October 10, 2007 Rev.79 by DTB
October 26, 2007 Rev.80 by DTB
October 31, 2007 Rev.81 by GJH/DTB convert to MSWord-based PDF
February 26, 2008 Rev.82 by GJH
April 9, 2008 Rev.83 by GJH
December 8, 2008 Rev.84 by GJH
July 6, 2009 Rec.85 by GJH/DTB Version 2.1
..........................................................................
The list of *references* used in this documentation is:
Adler, R.F., G.J. Huffman, A. Chang, R. Ferraro, P. Xie, J. Janowiak, B. Rudolf, U. Schneider,
S. Curtis, D. Bolvin, A. Gruber, J. Susskind, P. Arkin, E.J. Nelkin, 2003: The Version 2.1
Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979 Present).
J. Hydrometeor., 4(6), 1147-1167.
Adler, R.F., G.J. Huffman, P.R. Keehn 1994: Global rain estimates from microwave-adjusted
geosynchronous IR data. Remote Sens. Rev., 11, 125-152.
Arkin, P.A., and B. N. Meisner, 1987: The relationship between large-scale convective rainfall
and cold cloud over the Western Hemisphere during 1982-1984. Mon. Wea. Rev., 115, 51-74.
Grody, N.C., 1991: Classification of snow cover and precipitation using the Special Sensor
Microwave/Imager (SSM/I). J. Geophys. Res., 96, 7423-7435.
Hollinger, J.P., J.L. Pierce, G.A. Poe, 1990: SSM/I instrument evaluation. IEEE Trans. Geosci.
Remote Sens., 28, 781-790.
Huffman, G.J., 1997a: Estimates of root-mean-square random error contained in finite sets of
estimated precipitation. J. Appl. Meteor., 36, 1191-1201.
__________, ed., 1997b: The Global Precipitation Climatology Project monthly mean
precipitation data set. WMO/TD No. 808, WMO, Geneva, Switzerland. 37pp.
__________, R.F. Adler, D.T. Bolvin, G. Gu, 2009: Improving the global precipitation record:
GPCP Version 2.1. Geophys. Res. Lett., submitted.
__________,
R.F. Adler, B. Rudolf, U. Schneider, P.R. Keehn, 1995: Global precipitation
estimates based on a technique for combining satellite-based estimates, rain gauge analysis,
and NWP model precipitation information. J. Climate, 8, 1284-1295.
GPCP V2.1
__________, __________, P.A. Arkin, A. Chang, R. Ferraro, A. Gruber, J. Janowiak, R.J. Joyce, Janowiak, R.J. Joyce,
A. McNab, B. Rudolf, U. Schneider, P. Xie, 1997: The Global Precipitation Climatology
Project (GPCP) combined precipitation data set. Bull. Amer. Meteor. Soc., 78, 5-20.
Janowiak, J.E., P.A. Arkin, 1991: Rainfall variations in the tropics during 1986-1989. J.
Geophys. Res., 96, 3359-3373.
Joyce, R.J., J.E. Janowiak, G.J. Huffman, 2001: Latitudinal and seasonal dependent Zenith
Angle Corrections for geostationary satellite IR Brightness Temperatures. J. Appl. Meteor.,
40, 689-730.
Krajewski, W.F., G.J. Ciach, J.R. McCollum, C. Bacotiu, 2000: Initial validation of the Global
Precipitation Climatology Project over the United States. J. Appl. Meteor., 39, 1071-1087.
Legates, D.R, 1987: A climatology of global precipitation. Pub. in Climatol., 40, U. of
Delaware.
McNab, A., 1995: Surface Reference Data Center Product Guide. National Climatic Data Center,
Asheville,NC, 10 pp.
Morrissey, M.L., J. S. Green, 1991: The Pacific Atoll Raingauge Data Set. Planetary Geosci.
Div. Contrib. 648, Univ. of Hawaii, Honolulu, HI, 45 pp.
Rudolf, B., 1993: Management and analysis of precipitation on a routine basis. Proc. Internat.
WMO/IAHS/ETH SYMP. on Precip. and Evap., Slovak Hydromet. Inst., Bratislava, Sept.
1993, 1, 69-76.
Schneider, U., T. Fuchs, A. Meyer-Christoffer, B. Rudolf, 2008: Global precipitation analysis
products of the GPCC. GPCC Internet Publikation, DWD, 12 pp. Available online at
http://gpcc.dwd.de.
Susskind, J., J. Pfaendtner, 1989: Impact of interactive physical retrievals on NWP. Report on
the Joint ECMWF/EUMETSAT Workshop on the Use of Satellite Data in Operational
Weather Prediction: 1989-1993, Vol. 1, T. Hollingsworth, Ed., ECMWF, Shinfield Park,
Reading RG2 9AV, U.K., 245-270.
Susskind, J., P. Piraino, L. Rokke, L. Iredell, A. Mehta, 1997: Characteristics of the TOVS
Pathfinder Path A Dataset. Bull. Amer. Meteor. Soc., 78, 1449-1472.
Weng, F., N.C. Grody, 1994: Retrieval of cloud liquid water using the Special Sensor
Microwave Imager (SSM/I). J. Geophys. Res., 99, 25535-25551.
Wilheit, T., A. Chang, L. Chiu, 1991: Retrieval of monthly rainfall indices from microwave
radiometric measurements using probability distribution function. J. Atmos. Ocean. Tech., 8,
118-136.
Willmott, C.J., C.M. Rowe, W.D. Philpot, 1985: Small-scale climate maps: A sensitivity analysis
of some common assumptions associated with grid-point interpolation and contouring. Amer.
Cartographer, 12, 5-16.
WCRP, 1986: Report of the workshop on global large scale precipitation data sets for the World
Climate Research Programme. WCP-111, WMO/TD -No. 94, WMO, Geneva, 45 pp.
Xie, P., J.E. Janowiak, P.A. Arkin, 2000: An improved global precipitation index based on
satellite-observed outgoing longwave radiation. (to be submitted to J. Climate)
Xie, P., P.A. Arkin, 1998: Global monthly precipitation estimates from satellite-observed
outgoing longwave radiation. J. Climate, 11, 137-164.
__________, __________, 1997: Global precipitation: A 17-year monthly analysis based on
gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor.
Soc., 78, 2539-2558.
GPCP V2.1
__________, __________, 1996: Analysis of global monthly precipitation using gauge
observations, satellite estimates, and numerical model predictions. J. Climate, 9, 840-858.
...........................................................................
, __________, 1996: Analysis of global monthly precipitation using gauge
observations, satellite estimates, and numerical model predictions. J. Climate, 9, 840-858.
...........................................................................
A*citation list* that details refereed papers citing the GPCP is maintained on-line at
ftp://precip.gsfc.nasa.gov/pub/gpcp-v2.1/doc/gpcp_citation_list.pdf .
..........................................................................
14. Inventories
The *data set inventory* may be obtained by accessing the home pages or contacting the
representatives listed in section 12.
..........................................................................
15. How to Order Data and Obtain Information about the Data
Users interested in *obtaining data* should access the home pages or contact the representatives
listed in section 12.
..........................................................................
The *data access policy* is "freely available" with three common-sense caveats:
1.
The data set source should be acknowledged when the data are used. [One possible wording
is: "The GPCP combined precipitation data were developed and computed by the
NASA/Goddard Space Flight Center's Laboratory for Atmospheres as a contribution to the
GEWEX Global Precipitation Climatology Project."]
2.
New users should obtain their own current, clean copy, rather than taking a version from a
third party that might be damaged or out of date. Current users should check for updates and
new versions to avoid reliance on out-of-date data.
3.
Errors and difficulties in the dataset should be reported to the dataset creators.
4.
The GPCP datasets are developed and maintained with international cooperation and are used
by the worldwide scientific community. To better understand the evolving requirements
across the GPCP user community and to increase the utility of the GPCP product suite, the
dataset producers request that a citation be provided for each publication that uses the GPCP
products. Please email the citation to george.j.huffman@nasa.gov or
david.t.bolvin@nasa.gov. Your help and cooperation will provide valuable information for
making future enhancements to the GPCP product suite.
..........................................................................
GPCP V2.1