Multi-Satellite Remote Sensing of Global Water Surfaces Dynamics 1993-2005
Papa, F1; Prigent, C2; Rossow, W1
Inundations and wetlands are an integral part of the biochemical and the hydrological
cycles. First, they are the world's dominant natural source of methane (CH4), the only
one dominated by climate variations and the fluctuations in inundation and wetlands
appear to represent the major contributors to interannual variability in methane surfaces
emissions. Second, wetlands and inundated surfaces play a key role in global and local hydrological cycle. They regulate river hydrology and a part of the fresh water input to oceans by river discharge is processed through them. Flooded areas and the rivers are also sensitive to regional and global climate variations and therefore may be a good indicator of climate change.
However, water surfaces databases still suffer of a lack of reliable information at global scale over long time period. In this study, we present the first global estimate of monthly water surface extent and dynamic at 25km resolution for the 1993-2005 period.
The technique it self is based on the detection of inundations using the passive
microwave land-surfaces emissivities, estimated from SSM/I and ISCCP observations
and the use of ERS scatterometer and AVHRR visible and near infrared reflectances to
estimate the vegetation contribution to the passive microwave signal.
Results evaluation is made using independent datasets such as land surfaces databases,
rain fall rate from GPCP products, other satellites observations such as SAR or radar
altimeter and in-situ snow and runoff parameters. The satellite-derived water surfaces
estimates, comprising natural wetlands, inundation, irrigated rice culture and lake/river are relevant when compare to static wetlands and inundations observations databases. They also show a good agreement with the SAR-derived water surfaces extent over the Amazon River. Over the globe, the results show
a high correlation with the GPCP rain data product over the 13 years of study. The
correlations over specific regions are of a special interest as they provide valuable information on inundations
related to snow-melt or rain in upstream locations or inundations fed with direct precipitation at
the location. When focusing around the large tropical or boreal river basins, the analysis based on the comparison with water river levels derived from altimeter observations show a similar seasonal and inter-annual cycle over the 13 years. Over specific regions such as the large Arctic watershed, the seasonal and inter-annual variability of inundation extent is also closely link to the winter snow water equivalent and spring discharge variations. Preliminary analysis of the inter-annual variability of the dataset exhibits very interesting features that tend to show the sensitivity of the inundation extent to climate change. The water surfaces database is now used for hydrology modeling, methane studies in GCMs, estimations of water volume change in large river basin or as a proxy to study ENSO or monsoon regimes.