Semi-seasonal groundwater forecast using multiple data-driven models in an irrigated cropland

Francisco Munoz-Arriola,Alessandro Amaranto,Dimitri P Solomatine,Gerald Corzo,George Meyer

doi:10.2166/hydro.2018.002

Abstract

Abstract In agricultural areas where groundwater is the main water supply for irrigation, forecasts of the water table are key to optimal water management. However, water management can be constrained by semiseasonal to seasonal forecast. The objective is to create an ensemble of water table one- to five-month lead-time forecasts based on multiple data-driven models (DDMs). We hypothesize that data-driven modeling capabilities (e.g., random forests, support vector machines, artificial neural networks (ANNs), extreme learning machines, and genetic programming) will improve naïve and autoregressive simulations of groundwater tables. An input variable selection method was used to carry the maximum information in the input (precipitation, crop water demand, changes in groundwater table, snowmelt, and evapotranspiration) and output relationship for the DDMs. Five DDMs were implemented to forecast groundwater tables in an unconfined aquifer in the Northern High Plains (Nebraska, USA). Root mean squared error (RMSE) and Nash–Sutcliffe efficiency index were used to evaluate the skill of the model and three hydrologic regimes were determined statistically as high, mid, and low groundwater table levels. Additionally, varying storage regimes were used to construct rising and falling limbs in the tested well. Results show that all models outperform the baseline for all the lead times, ANNs being the best of all.

Highlights

Groundwater is a key resource to sustaining hydrological conditions of a watershed as well as agricultural activities
The contribution of SNM to forecast GW level changes is minimized due to the absence of snow storage in the initialization time produced by constrained input variable selection (CIVS)
This might be attributed to the contribution of snowmelt to recharge as lead time reaches five months (July and August) when groundwater withdrawals are at their top, so crop irrigation requirements can be supplied

Summary

Introduction

Groundwater is a key resource to sustaining hydrological conditions of a watershed as well as agricultural activities. Reliable water supply policies require accurate estimations of current and future water table depths and their fluctuations (see, for example, Coulibaly et al ). For this purpose, physically based, statistical, and data-driven modeling techniques are widely used. Hanson et al ( ) implemented the physically based model MODFLOW and the farm process package (MFFMP), parameterizing the micro- and macro-scale crop irrigation requirements or evapotranspirative needs to simulate the conjunctive use of surface water and groundwater. In agricultural areas where GW-based irrigation is used to satisfy evapotranspirative needs, variables such as evapotranspiration, crop water demand, precipitation and groundwater well levels could integrate the imbedded complexity of aquifer recharge with respect to streamflow, precipitation fluctuations, and well management

Objectives

Methods

Results

Conclusion