| |
The aggregated 17-class map provides
broad categories of irrigated agriculture, rainfed
agriculture, and few other classes of importance. Often,
many users would just need such broad classes. The
disaggregated classes provide discrete classes and are
often invaluable for detailed LULC assessments at
Global, regional, National, and local levels. The class
labeling in disaggregated classes of GMLULCA are only
indicative and can be improved based on further
knowledge about the class/es in consideration.
Unlike the GMIA and GMRCA, no attempt
was made to calculate sub-pixel areas (SPAs) for
GMLULCA. Like various other studies of LULC, discussed
soon after Table 1 below, GMLULCA provides full pixel
areas (FPAs)-see Table 1. In comparison to USGS 1-km
1992-93 AVHRR data (see Loveland et al., 1999, and
Loveland et al. 2000) based study, the IWMI LULC results
using SPOT 1-km and AVHRR 10-km data at the end of the
last millennium showed (Table 1): (a) croplands
increased by nearly 14 percent, (b) shrublands,
grasslands, and savannas decreased by nearly 16 percent,
and (c) forests increased by nearly 6.2 percent. No
attempt was made to calculate areas for different
seasons. The GMLULCA FPAs are determined for
June-October which is the major cropping season of the
World.
The first 8 classes in aggregated
GMLULCA 17-class map were irrigated areas (class 1-5
pure irrigated, class 6-7 supplemental, and class 8 LULC
with irrigation fragments). The full pixel areas (FPAs)
of these classes during June-October were: 537.1 Million
hectares for class 1-5, 319.8 million hectares for class
7-8, and 343.7 million hectares for class 8. This sums
up to about 1.2 billion hectares of FPA irrigated. In
reality, through sub-pixel de-composition technique (see
GMIA page and area calculation pages in this web site),
we established actual irrigated area to be 318 million
hectares. Thereby, on an average, only 26.5 %
(0.318/1.2) of the pixel is actually irrigated. The
exact fraction will vary from class to class, with
classes such as class 1 or 2 having significantly higher
fraction (SPA) of irrigated compared to classes 6-8.
The full pixel areas (FPAs) of IWMI
GMLULCA at the end of the last millennium versus USGS
LULC for 1992-93 for 3 key classes are shown in Table
1.
In the past, the most commonly used
global land use/land cover (LULC) datasets were
primarily non-remote sensing based, produced using data
from various maps at 100-km grid by Matthews (1983), and
50-km grid by Olson and Watts (1982), Olson (1994), and
Wilson and Henderson-Sellers (1985). More recently,
AVHRR and MODIS sensor data have been widely used to
produce global LULC. The 1992-93 AVHRR 1-km data were
used by USGS (see Loveland et al., 1999, Loveland et al.
2000) and University of Maryland (DeFries et al. 1995,
DeFries et al. 1998) to produce Global LULC datasets.
These data also were used by IGBP (see see Loveland et
al., 1999, Loveland et al. 2000). The most recent LULC
products are from Boston University using MODIS (Friedl
et al., 2002, Zhan et al., 2000). It is well known that
no two global datasets match (DeFries and
Townshend,1994) as a result of differences in methods,
data sources, data types, data calibration, and data
acquisition modes. Also, class definitions and how
classes are aggregated to meet user requirements or
target groups. The disaggregated 75-class IWMI GMLULCA
can be used to derive classes that match the class names
of USGS and MODIS LULC classes. The particular strength
of IWMI’s GMLULCA will be in its emphasis on irrigated
and rainfed croplands and in establishing their
sub-pixel areas (SPAs).
References
DeFries,
R., Hansen, M., and Townshend, J., 1995, Global
discrimination of land cover types from metrics derived
from AVHRR Pathfinder data. Remote Sensing of
Environment, 54, 209–222.
DeFries, R., Hansen, M., Townsend, J. G. R., &
Sohlberg, R. (1998).Global land cover classifications at
8 km resolution: the use of training data derived from
Landsat imagery in decision tree classifiers.
International Journal of Remote Sensing, 19, 3141–
3168.
Friedl, M.A., McIver, D.K., Hodges, J.C.F., Zhang,
X.Y., Muchoney, D., Strahler, A.H., Woodcock, C.E.,
Gopal, S., Schneider, A., Cooper, A., Baccini, A., Gao,
F., Schaaf , C. 2002: Global land cover mapping from
MODIS: algorithms and early results, Remote Sensing of
Environment, Vol. 83 (1-2), pp. 287-302.
Loveland,
T.R., Reed, B.C., Brown, J.F., Ohlen, D.O., Zhu, Z.,
Yang, L., Merchant, J.W. 2000. Development of a global
land cover characteristics database and IGBP DISCover
from 1 km AVHRR data. int. j. remote sensing, 21 (6
& 7): 1303–1330.
Matthews, E., 1983, Global
vegetation and land use: new high resolution data bases
for
limited studies. Journal of Climatology and
Applied Meteorology, 22, 474–487.
Muchoney, D., Strahler, A., Hodges, J., &
LoCastro, J. (1999). The IGBP DISCover confidence sites
and the system for terrestrial ecosystem
parameterization: tools for validating global land-cover
data. Photogrammetric Engineering and Remote Sensing,
65(9), 1061– 1067.
Olson, J. S., 1994, Global Ecosystems Framework:De.
nitions. Internal Report, USGS EROS Data Center, Sioux
Falls, SD, USA.
Olson, J.S., and Watts, J.A., 1982. Major World
Ecosystem Complex map, Oak Ridge National Laboratory,
Oak Ridge, Tennessee.
Zhan, X., DeFries, R., Townshend. J. R. G, Dimiceli,
C., Hansen, M., Huang, C., and Sohlberg, R. 2000. The
250m global land cover change product from the Moderate
Resolution Imaging Spectroradiometer of NASA’s Earth
Observing System. int. j. remote sensing, 21 (6 &
7): 1433–1460 |