Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2021 → Seasonal predictions : from global to regional information for decision support in water resources management in semi-arid regions

Universität Augsburg (2021)

Seasonal predictions : from global to regional information for decision support in water resources management in semi-arid regions

Tanja Christina Portele

Titre : Seasonal predictions : from global to regional information for decision support in water resources management in semi-arid regions

Auteur : Tanja Christina Portele

Université de soutenance : Universität Augsburg

Grade : Doktorgrades 2021

Résumé partiel
The increasing tension within the water-energy-food nexus in face of climate change requires the sustainable use and management of available water resources. This is particularly relevant in semi-arid regions of South America, Africa and West Asia : already, these regions are facing high precipitation variability, and are additionally exposed to high population growth rates, politically induced unsustainable use of water resources, and transboundary water management conflicts. It is shown that for the climate-sensitive regions of Northeast Brazil, Iran, West Africa, and Northeast Africa, the relative frequency of droughts has increased significantly from 10 to 30 % in recent decades. This calls for proactive measures to aid climate proofing and mitigate climate risks. For water allocation during drought or for flood control, improved knowledge of the upcoming rainy season in semi-arid regions can be the basis for a more robust and sustainable water management. Here, seasonal forecasts with forecast horizons up to seven months ahead can offer great opportunities to support regional water management. State-of-the-art seasonal forecasting systems already reach resolutions that are suitable for regional applications, e.g., the latest seasonal forecasting system version 5 (SEAS5) of the European Centre for Medium-Range Weather Forecasts (ECMWF) with a horizontal resolution of 36 km. Indeed, with skillful and reliable forecasts for the coming months, decision-makers in the water, agriculture, and energy sectors could induce a more timely, proactive and sustainable reservoir management and seed selection, thereby reducing damage and loss. Decision-makers, however, often still hesitate to use seasonal forecasts claiming their lack of reliability and the inherent uncertainty due to their probabilistic nature. In many cases, statistical performance measures for forecast quality cannot provide actual decision support. Therefore, the potential economic value (PEV) is implemented demonstrating the possible relative savings when basing decisions for preventative water management action on seasonal forecasts compared to optimal early action. This measure of forecast value therewith demonstrates the economic benefit of integrating seasonal forecasts in decisions incorporating the user’s cost-loss situation and the fore- cast probability. The methodology used in this study allows the use of unprocessed global seasonal forecasts for regional analysis. It is shown that a proactive approach based on seasonal forecasts in the case of droughts can achieve potential economic savings of up to 70 % of those from optimal early action. For very warm months and droughts, savings of at least 20 % are achieved even with forecast horizons of up to seven months ahead. For one particular large dam in Sudan, the Upper-Atbara Dam, the savings are explicitly quantified in monetary terms if the dam was operated with early drought mitigation. In fact, avoidable losses of 16 Mio US$ are revealed in one example year.

Mots clés  : Seasonal forecast ; Statistical downscaling ; Dynamical downscaling ; Economic value ; Spatial patterns

Présentation

Version intégrale (37,4 Mb)

Page publiée le 1er novembre 2022