The
Generic IWMI 951 Class
map (Generic-IWMI-951) is the
base product from which the Global irrigated area
map (GIAM), the Global Map of Rainfed Cropland
Areas (GMRCA), and Global Map of Land Use/Land
Cover Areas (GMLULCA) were derived. The
methodology on how the 951 class map was produced
has been presented in the detailed documentation presented
in this paper. First, the irrigated area classes,
the rainfed area classes, and the other land
use/land cover (LULC) classes were identified and
labeled based on an elaborate set of methods and
techniques (see documentation) for each
of the seven segments of the World. The seven
segments were:
1.
Precipitation less than 360 mm per year
segment (PLT360-segment);
2.
Precipitation greater than 2400 mm per year
segment (PGT2400-segment);
3.
Temperature less than 280 degree Kelvin
segment (TLT280-segment);
4.
Forest cover greater than 75 mm segment
(FGT75-segment);
5.
Elevation greater than 1500 mm segment
(EGT1500-segment);
6.
All other areas of the World segment
(AOAW-segment); and
7.
Rainforest synthetic aperture radar segment
(RFSAR-segment).
The
segments were masked and analyzed using the
time-series mega-file data from SPOT 1-km, AVHRR
10-km, and secondary datasets such as 1-kmGTOPO30
elevation, 1-km forest cover, 40-yr monthly
rainfall, and 10-km AVHRR skin temperature. The
classes identified from each of the segments were
mosaicked into a single file that added up to
951 classes and is
called Generic IWMI 951 Class Map
(Generic-IWMI-951). The 951
classes were refined into various aggregated and
disaggregated maps specifically for GIAM, GMRCA, and GMLULCA.
The
Generic 951 class map and image are specifically
useful for those interested in producing their own
refined map. For example, a more detailed
irrigated area map or rainfed cropland map, or a
LULC map of specific Countries or regions will be
of value when the can better combine and/or label
classes based on local knowledge. It must be noted
that the class labels in the 951 class map/image
are indicative rather than definitive. So a user
can always label a class better. The
characteristics include, for each of the 951
classes: (a) month-by-month 10-km AVHRR spectral
band reflectivity for band 1 (red) and band 2
(near-infrared), (b) month-by-month 10-km AVHRR
brightness temperature for thermal bands 4 and 5;
(c) month-by-month 10-km AVHRR NDVI, (d)
month-by-month 1-km SPOT vegetation NDVI, (e)
elevation (meters), (f) skin temperature (degree
kelvin), (g) mean rainfall (mm/month), and (h)
tree cover (percent).