Seasonal forecasting of dam water resources using optimized hybrid models under unprecedented drought conditions

Domaine:

climateenvironment and energy

Type de record:

paper

Créateur:

HanHadBouKha

Éditeur:

UniMohCenUni

Éditeur:

CCSD

Hôte:

International audience In the Oum Er Rbia watershed, Morocco, dam water resources play a crucial role in prolonged drought conditions, particularly in the case of the Al Massira Dam, which has been a strategic reservoir for drought resilience since its inauguration. In this study, optimized pipelines of explainable artificial intelligence (XAI) models were developed for monthly forecasts of water resource variations at the Al Massira dam, which has been affected by unprecedented drought since 2019. The architectures of the models developed incorporate Bayesian optimization via Optuna for identifying the best hyperparameters, advanced feature selection methods, and lagged regressors of teleconnection indices, drought indices, and hydroclimatic variables. The performance of the models was first evaluated in terms of their ability to forecast dam water volume up to 6 months ahead under near-normal hydroclimate conditions. Next, model performance was assessed under a scenario of unusual changes in time series. The Bayesian probabilistic LSTM (ProbLSTM) reached the maximum predictive skill score (Skill=86.2 %, NMAE=3.6 %), followed by the generalized additive model (GAM) (Skill=85.3 % and NMAE=3.4 %).However, the light gradient boosting machine (LightGBM) showed high uncertainty when forecasting water volumes under unusual drought conditions, with Skill= 75.1 % and NMAE= 11.2 %. Thus, from an operational perspective for seasonal forecasting, ProbLSTM and the GAM are preferable because of their low performance variability under a scenario of unusual changes in time series and their high predictive performance.

Visit

hal.science

Similaires

Probabilistic Seasonal Forecasting of African Drought by Dynamical ModelsDownscaling GRACE total water storage data using random forest: a three-round validation approach under drought conditionsSCREENING OF NORTH AFRICAN BARLEY GENOTYPES FOR DROUGHT TOLERANCE BASED ON YIELDS USING TOLERANCE INDICES UNDER WATER DEFICIT CONDITIONSShort-Term and Medium-Term Drought Forecasting Using Generalized Additive ModelsForecasting HIV Incidence in Somalia Using Single and Hybrid Time Series Models: A Comparative Analysis HIV Incidence Forecasting in Somalia Using Time Series and Hybrid ModelsStochastic Drought Forecasting Exploration for Water Resources Management in the Upper Tana River Basin, Kenya

Probabilistic Seasonal Forecasting of African Drought by Dynamical Models

Abstract As a natural phenomenon, drought can have devastating impacts on local pop

Downscaling GRACE total water storage data using random forest: a three-round validation approach under drought conditions

International audience The application of GRACE satellite-derived Total Water Storage

SCREENING OF NORTH AFRICAN BARLEY GENOTYPES FOR DROUGHT TOLERANCE BASED ON YIELDS USING TOLERANCE INDICES UNDER WATER DEFICIT CONDITIONS

Water deficit is one of the most constraining factors for the growth, development and yields of plan

Short-Term and Medium-Term Drought Forecasting Using Generalized Additive Models

Forecasting extreme hydrological events is critical for drought risk and efficient water resource ma

Forecasting HIV Incidence in Somalia Using Single and Hybrid Time Series Models: A Comparative Analysis HIV Incidence Forecasting in Somalia Using Time Series and Hybrid Models

This upload contains the manuscript and supporting materials for a time series forecasting study of

Stochastic Drought Forecasting Exploration for Water Resources Management in the Upper Tana River Basin, Kenya

This chapter presents the trend of drought as a stochastic natural disaster influenced by the climat