Application of inclusive multiple model for the prediction of saffron water footprint

Applying new approaches in the management of water resources is a vital issue, especially arid and semi-arid regions. The footprint key index management. Therefore, it necessary to predict its changes for future durations. soft computing model one most widely used models predicting estimating agroclimatic variables. purpose this study green blue footprints saffron product using model. In order select effective variables prediction footprints, individual input was eliminated by effect each on residual mean square error (RMSE) measured. first stage, Group Method Data Handling (GMDH) evolutionary algorithms have been applied. next output incorporated into Inclusive Multiple Model (IMM) as three homogenous Finally, uncertainty caused parameters evaluated. contributions research are introducing optimized GMDH ensemble BWF, GWF, analysis investigating inputs GWF BWF. results indicated that important affecting plant transpiration, evapotranspiration, yield, since removing these significantly increased RMSE (range=11–25). Among models, best performance belonged NMRA (Naked Mole Ranked Algorithm) due fast convergence high accuracy outputs. regard, IMM has better (FSD=0.76, NSE=0.95, MAE) = 8, PBIAS= 8) than alternatives applying outputs several lowest based (p 0.98, r 0.08).