|
This is the version 2.0 release (update; as of May 10,
2007) of the International Water Management Institute’s
(IWMI’s) Global irrigated area map (GIAM) and associated
products and data.
The GIAM products are produced using time-series data of:
(a) AVHRR 10-km monthly from 1997-1999, (b) SPOT 1-km monthly
for 1999, (c) GTOPO30 1-km elevation, (d) CRU 50-km grid
monthly precipitation from 1961-2000, (e) AVHRR derived 1-km
forest cover, and (f) AVHRR 10-km skin temperature. In
addition JERS SAR data was used for the African and South
American rainforests.
The front-page provides gateway to 4 product lines. The
primary product, and overwhelming emphasis, is on Global Irrigated
Area Map (GIAM). You can click on appropriate
places in this front-page that will lead you to the above
product line.
There are many unique features in the IWMI’s GIAM product
line. First, this is the very first satellite sensor based
global irrigated area map. Second, the resolution of the map
(10-km) is the best that is presently available for irrigated
areas at global level. Third, the area calculations are done
for each season. So the area irrigated @ the end of the last
millennium for the entire world was: (a) 257 Mha during
June-September, (b) 174 Mha during October-February, and (c)
41 Mha during March-May. Further, there is a flexibility to
calculate areas every month. Fourth, this is NOT just a map.
There are suite of products that consists of maps, images,
class characteristics, area calculations, snap-shots and
photos, animations, and accuracies. There are numerous
advantages of such a product line. For example, disaggregated
class images can be downloaded and a more refined map can be
created with local expertise for one’s area of interest. The
irrigated areas are used to create a 20-year animations using
AVHRR monthly time-series, so that one can spatially re-create
the history of an irrigated area class. The class
characteristics facilitate deriving crop calendar,
sowing-peak-harvest dates of each class, and determine whether
a class is single, double, or continuous crop. Fifth, the
study develops and/or adopts a suite of innovative methods and
techniques to map irrigated areas of the World at Global to
local levels and at all resolutions or scale. The methods
include spectral matching techniques (SMTs), image
segmentation, decision tree algorithms and spatial modeling,
data fusion, space-time spiral curves,
brightness-greenness-wetness 2-dimensional feature space
plots, NDVI time series plots, NDVI thresholds, principal
component analysis, and unsupervised clustering algorithms.
The wide array of ground truth data was also used. This
included ground truth data of the Indus-Ganges river basins,
Krishna river basin, IWMI’s ground truth data of the World
that included data for Middle East and Africa, the degree
confluence project data of the world, and the 150-m Landsat
geocover mosaic of the world. |
|