Employment of the SSEB and CROPWAT Models to Estimate the Water Footprint of Potato Grown in Hyper-arid Regions of Saudi Arabia
1K. Al-Gaadi, 1R. Madugundu, 1E. Tola
1. King Saud University
2. Precision Agriculture Research Chair, King Saud University,
Quantifying crops’ water footprint (WF) is essential for sustainable agriculture especially in arid regions, which suffers from harsh environmental conditions and severe shortage of freshwater resources such as Saudi Arabia. In this study, WF of irrigated potato crop was estimated for the implementation of precision agriculture techniques. The CROPWAT and the Simplified Surface Energy Balance (SSEB) approaches were adopted. Soil, plant, and yield samples were randomly collected from six potato fields belongs to the Saudi Agricultural Development Company, Wadi-Ad-Dawasir region, Saudi Arabia. Subsequently analyzed for potato tuber yield (t ha-1). The consumptive crop water use (CWU) was computed, as the actual evapotranspiration (ETa), using the SSEB algorithm. The vegetation indices (NDVI, normalized difference RedEdge-NDRE, MSAVI, RedEdge chlorophyll index-RECI and NDMI) were computed from the obtained sentinel-2 and Landsat-8 data and used as inputs to predict the crop productivity (CP), the CWU, and subsequently the WF. The results indicated that the NDRE showed the best prediction accuracy for potato CP (R2 = 0.72, P>F = 0.021) followed by the MSAVI (R2 = 0.64, P>F = 0.018). The CWU, however, was successfully estimated (as ETa) using the SSEB algorism with an overall accuracy of 89.2%, where the differences between the actual amounts of irrigation water and the estimated ETa ranged between 12.6% (autumn) and 10.6% (winter) during the season. Based on the CROPWAT-SSEB estimates, the average total WF of potato was found to be 6846 m3 ton-1. Out of this, the green and blue WF contribution was estimated at averages of 8% and 92%, respectively. Comparison between the blue WF from the SSEB-CROPWAT and the field-data based estimates showed a good agreement (nRMSE = 8.4%, nMBE = 12.9% and relative error –RE ranging from 1.1 to 14%. The effect of planting date on the WF estimation also studied in this research and slight variation of 1.5% (prior) to 7.5% (after) about two months to the baseline planting dates was noticed. It can be concluded that the WF assessment could be satisfactorily estimated using the CROPWAT/SSEB models for irrigation management.
