Practical analysis of remote sensing estimations of water use for major crops throughout the Urmia Lake basin

Abstract

Remote sensing techniques are used to estimate the high spatial and temporal hydrological variable of evapotranspiration. This is certainly considered a step forward, compared to alternate point measurements. It is important, however, to properly interpret these estimations to practical information for water managers. In this study, results of two separate remote sensing studies (FAO-IHE and RSRC) on estimation of actual evapotranspiration for major crops in the Urmia Lake basin were compared with widely accepted theoretical estimations of corresponding Irrigation Requirements (IR) using CROPWAT. CROPWAT estimations represent the irrigation requirements of crops under no stress, whereas RS techniques are used to estimate actual crop water use which may be under nonideal growing situations. Comparison of irrigation requirements and actual water use for different crops at various locations within the Urmia Lake basin can provide practical information on the status of field water managements, i.e., sufficiency of applied irrigation. Such analysis provides grounds for possible improvement on water management in a highly competitive water used region. The RS crop water estimations results were crosschecked with CROPWAT estimations of irrigation requirements, as the reference range, using ground data adopted from 21 scattered synoptic weather stations across the basin, representing the widely variable crop-water demand in the region and corresponding effective precipitations. FAO-IHE used the SEBAL and RSRC used the METRIC algorithm to produce ETa maps. The available land-use maps with 30-m by 30-m pixels were used for designation of the type of field crops. RSRC maps were produced using 1-km by 1-km daily satellite images. At this scale, a mixed combination of crops occurs in one pixel. To provide a more realistic comparison, the relative area of each crop cited from the land use maps was used to compare the ETa estimates by RSRC with the weighted average of irrigation requirements. The results showed the actual water use estimated from remote sensing images were generally higher than Irrigation Requirements. The difference can be attributed to overirrigation which is a common practice in the region. RS estimated actual crop-water use in rainfed lands was also compared with temporal local precipitations. In some areas, actual crop-water use in rainfed lands was more than the corresponding effective precipitation estimates. This observation cannot practically be justified and can be drawback in the preparation of land-use maps. Practical conclusions were drawn from comparison of actual crop-water use and theoretical estimates. Locations of possible over-/under-irrigation were identified, for which improved water management can provide means for saving water for the restoration of the desiccating Urmia Lake. The feasibility and versatility of RS techniques have motivated researchers to increase adoption of this method, provided that the uncertainties can be resolved.