Thesis (Ph.D.) - Cairo University - Faculty of Agriculture - Department of Agricultural Engineering

Egypt suffers from water scarcity due to the increase in the population, climate change, and the lack of integrated management of water resources. Therefore, accurate evaluation of irrigation water needs for crops is urgent to achieve water management sustainability. Water footprint considers an indicator of water management sustainability. So this study investigated the impact of climate change on potato yield and water footprint in 10 governorates in the Nile Delta, Egypt during the period from 1990 to 2016. Based on the results of the BWF calculation, Three governorates were selected (Al-Gharbia, Al-Dakahlia, Al-Beheira) to develop and compare between four machine learning (SVM, RF, XGB and ANN). To select the best model in the best scenario, which achieve a high degree of accuracy and low error for predicting blue WF of potato. The results showed that, the spatial distribution of climate parameters shows that the highest precipitation was reported in Alexandria followed by Kafr El-Sheikh during winter season. On contrast, the maximum ETC was in the south part followed by the middle governorates and the lowest located in the northern governorates. The potato water footprint in Delta Egypt decreased from 170 m3 ton-1in 1990 to 120 m3 ton-1 in 2016. The blue water footprint contributes more than 75% of the total water footprint, while the green water footprint contributes less than 25%. The XGB and ANN models generated good result in estimating WF through the testing stage with high accuracy more than 90% and less errors 0.25, R2 = 0.90, RMSE = 3.6 m3/t, NSE= very good SI = Fair in the three governorates. The results demonstrated that Sc.5 with the XGB and ANN model is good enough for assessing BWFP if only vapor pressure deficit, precipitation, solar radiation, crop coefficient data are available

Issued also as CD