. 2021 Feb 8;11:3295. doi: 10.1038/s41598-021-81233-4

Table 2.

Regional predictors of fire.

Region code	Region name	Monthly cor.	Interannual cor.	Spatial cor.	Burned area	BA trend	Anomaly cor.	Score
Region code	Region name	$r_{T}$	$r_{IA}$	$r_{S}$	BA	LT	$r_{An}$	P	Vars	Variables
NHAF	Northern Hemisphere Africa	0.87	0.77	0.88	151.9	− 1.60	0.53	88.1	3	pr, cld, pop
SHAF	Southern Hemisphere Africa	0.91	0.32	0.92	180.2	0.37	0.36	73.5	3	gppl1, ts, cld
SA	South America	0.92	0.89	0.85	30.7	− 0.17	0.91	94.8	7	gpp, gppm1s, pr, ts, cld, pop, rdtot
SA	South America	0.91	0.81	0.79	30.3	− 0.07	0.85	91.9	5	gpp, gppm1s, pr, ts, cld
SEAS	South and Southeast Asia	0.89	0.68	0.87	11.8	− 0.19	0.77	87.7	8	gpp, gppm1, pr, ts, cld, vp, pop, rdtot
SEAS	South and Southeast Asia	0.88	0.64	0.89	12.0	− 0.22	0.73	86.0	6	gpp, gppm1, pr, ts, cld, pop
TCAM	Temperate and Central America	0.77	0.84	0.74	6.3	− 0.01	0.84	90.7	5	gpp, gppl1, pr, ts, cld
TCAM	Temperate and Central America	0.77	0.78	0.73	6.0	0.01	0.79	89.1	4	gpp, gppl1, pr, ts
BONA	Boreal North America	0.83	0.56	0.67	3.3	0.00	0.57	81.2	6	gpp, gppl1, pr, ts, cld, vp
BONA	Boreal North America	0.83	0.55	0.56	3.8	− 0.01	0.57	79.4	3	gpp, gppl1, pr
AUS	Australia	0.86	0.91	0.91	50.2	0.21	0.92	95.1	6	gppm1s, gpp, gppl1, ts, cld, vp
AUS	Australia	0.86	0.90	0.88	50.5	0.03	0.90	94.6	3	gpp, gppl1, cld
CEAS	Central Asia	0.70	0.72	0.79	16.2	− 0.18	0.55	85.5	4	gppl1, pr, cld, vp
BOAS	Boreal Asia	0.65	0.59	0.82	9.5	0.01	0.62	82.2	4	gppm1, pr, ts, vp
EQAS	Equatorial Asia	0.80	0.93	0.87	2.0	0.02	0.95	95.0	3	pr, ts, cld
EQAS	Equatorial Asia	0.76	0.93	0.78	1.9	0.00	0.94	93.7	1	pr
EUME	Europe and Middle East	0.83	0.34	0.67	2.6	0.00	0.35	70.7	2	pr, cld

Performance of the best and minimal models for each region with respect to each of the five performance measures described in Methods, along with the aggregate performance score. In some regions, the best model is the same as the minimal model. Also mentioned are the variables that form the inputs of the models. BA is Burned Area, and LT is long-term trend in spatially aggregated yearly timeseries. Variables are as follows: gppl1—cumulative GPP, gppm1—growing season GPP (northern hemisphere), gppm1s—growing season GPP (southern hemisphere), pr—precipitation, ts—temperature, cld—cloud cover, vp—vapour pressure, rdtot—total road network density, pop—population density. All models include vegetation type fractions, including cropland fraction. The model for NHAF uses yearly vegetation fractions, whereas rest of the models use a single snapshot.