Characterization of salinity and sodicity in semi-arid irrigated agricultural lands using remote sensing

Abstract

Surface salinity processes are highly dynamic and the methods needed to properly detect them must respond to that dynamism making remote sensing a tool particularly well-suited for the management of salinized lands. It allows the monitoring of affected lands for the prevention of serious degradation through appropriate and timely action it is less costly in terms of time and resources than conventional methods and it is suited to the monitoring of large areas. Researchers are exploring how it can be adapted to the detection of moderate levels of salinity that could perhaps help to better prevent further degradation. The main aim of this research is to assess for the first time the potential of the ALI sensor (Advanced Land Imager) on board the EO-1 satellite, with its rich infrared bands, for the identification and mapping of salinity and sodicity. Through the testing of different salinity indices found in the literature, semi-empirical predictive models were developed which could be suited to the characterization and mapping of sodic and saline soil conditions in semi-arid agricultural areas, using the Tadla's irrigated perimeter of Morocco as a test case. Predictive models were based on a second order regression analysis calculated between the E.C. of soils affected by salinity and sodicity, and different spectral salinity indices using spectroradiometric ground measurements. Emphasis was placed on detecting slight and moderate soil salinity and sodicity, which has been considered a challenge in the past. Semi-empirical models were derived from the data, and applied to an ALI image for analysis. Visual comparisons and statistical validation of these models using ground truth were undertaken in order to identify the best model for the mapping of salinity and sodicity in the Tadla's irrigated perimeter of Morocco.