Exploratory Predictive Tool for Fire Activity in the Kalimantan Region, Indonesia

This page presents experimental forecasts of the likelihood of high or low fire activity in the provinces of South, Central, West, and East Kalimantan in Indonesia. This experimental research was conducted by IRI through a partnership with Bogor Agriculture University with support from CARE Indonesia. IRI is currently working with the Indonesian Meteorological Service (BMG) to compare methodologies for forecasting relevant to fires.

By monitoring the NINO4 index of the Sea Surface Temperature (SST) from April to September, it is possible to estimate the fire activity one to two months in advance. This analysis is based on data from 1998-2006 on fire hotspots (derived from NOAA-AVHRR) and on the NINO4 index. The four graphs that appear further down on this page show the results of this analysis for each of the four Kalimantan provinces in Indonesia.

The time series of the NINO4 index (shown in the graph below) is updated every month. Based on its relationship to fire activity, users may view the NINO4 data and forecast 1-2 months ahead whether fire activity is likely to be above or below median levels.

Monthly NINO4 Index, November 1998 to Present

Access to Data

Data: Monthly NINO4 index based upon Reynolds and Smith OIv2 sea surface temperature anomalies
Data Period: November 1998 - present
Base Period for Climatology: January 1971 - December 2000
Data Source: NCEP Environmental Modeling Center

Description: The time series above gives the value of the NINO4 sea surface temperature anomaly index in the central equatorial Pacific Ocean (description of NINO4 index) based upon sea surface temperature anomalies in the Reynolds and Smith OIv2 data set (description), with 1971-2000 as the climatological base period. NCEP is an acronym for the US National Centers for Environmental Prediction.

Predicting Fire Activity

Province:
Number of fires threshold:	NINO4 threshold:
Month lag:

Decision Matrix	NINO4 below median	NINO4 above median
Fires above median	7 (32%)	16 (70%)
Fires below median	15 (68%)	7 (30%)

Description

This graph shows the relationship between the number of fires and the NINO4 index in the previous month(s) for the Tengah Kalimantan province. The four colors mark the thresholds of interest chosen for NINO4 and the number of fires. To determine the risk of fire activity 1 month(s) later, use the value of the NINO4 index in May, June, July, August, or September, and find the estimated number of fires corresponding to that NINO4 level. Past and current values for the NINO4 index can be found through the graph above.

The decision matrix represents the number of points in each of the four quadrants defined by the color code. The decision matrix can be used to identify the probability of having above- or below- median fire values respective to the NINO4 index.

Data Sources
NINO4
Data: Monthly indices representative of the NINO 4 region
Data Source: National Centers for Environmental Prediction, Environmental Modeling Center, Climate Modeling Branch (NINO4)
Reference: Reynolds, R.W., N.A. Rayner, T.M. Smith, D.C. Stokes, and W. Wang, 2002: An Improved In Situ and Satellite SST Analysis for Climate. J. Climate, 15, 1609-1625

Fires
Data: Monthly number of fires in Kalimantan provinces, province-wide
Data Source: The data were acquired by the South Sumatra Forest Fire Management Project (SSFFMP) and the Japanese International Cooperation Agency (JICA)-in collaboration with the Ministry of Forestry of the Republic of Indonesia and collected by the Bogor Agricultural University National Centers for Environmental Prediction, Environmental Modeling Center, Climate Modeling Branch