Drought Identification through the Soil Moisture Estimated By Satellite and By AGCM-SSiB Model of CPTEC/INPE

The drought definition can change with the nature of water deficit and applications. Normally four categories of drought are used: (1) Meteorological drought - when atmospheric conditions cause low precipitation; (2) Hidrological drought - deficit in surface and subsurface water; (3) Agricultural drought - deficit of water for plants; (4) Socioeconomic drought - when human activities are affected.

The drought occurence depend of many elements: precipitation anomaly and/or temperature anomaly,anomaly of hidrological variables, terrestrial ecosystem condition and/or human activity. Therefore, drought monitoring  need to consider the total environmental moisture, using all indicators that is possible. The indexes more common are: PDSI - Palmer Drought Severity Index, SWPI - Surface Water Supply Index, SPI - Standardized Precipitation Index and Soil Moisture Percentiles.

Drought events on Brazil affect fundamental economy sectors and population supply,  therefore directly affect agricultural production, energy production and water supply. According with Brazilian Atlas of Natural Disasters, around 48% of occorence of natural disaster are drought events, when the region most affected is Northeast followed by South and Southeast.

Soil moisture is difficultly observated in large scale. To get around this problem is possible to use soil moisture estimated by satellite, mainly in microwave range of the electromagnetic spectrum. This estimate can detect soil moisture in the surface layer, around 5 cm. The satellite data product used in this work is development and provided by European Space Agency (ESA). The product is a combination of passive and active observations of many satellites.

Using the Standardized Precipitation Index (SPI)  that is a index widely used in drought monitoring centers, were selected 3 cases for analisys: Jan/1998 to Dec/1999, Mar/2007 to Dec/2008 and Jan/2012 to May/2013. Then, it was calculeted soil moisture percentiles for satellite estimates and land surface SSiB model. The land surface SSiB model is coupled with atmospheric general circulation model of CPTEC/INPE, in the version T062L24 that have around 200 km of horizontal resolution and simulation was in the period between Jan/1998 and Dec/2013.

Considering that the drought occurence is when the percentile of soil moisture is less than 20, the results indicate that satellite estimates although represent only the surface layer can identify similar drought regions to SPI and find that  have a time to response of precipitation occurence, increasing the drought of the hidrological viewpoint. Already  land surface SSiB model have difficult to reproduce the events continuously and when the event was simulated correctly, the drought region generally indicated percentiles higher than is observed.