Stochastic Simulation of Nonstationary Rainfall Fields, Accounting for Seasonality and Atmospheric Circulation Pattern Evolution
[EN] A model for generating daily spatial correlated rainfall fields suitable for evaluating the impacts of climate change on water resources is presented. The model, termed Stochastic Rainfall Generating Process, is designed to incorporate two major nonstationarities: changes in the frequencies of...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión enviada para evaluación y publicación |
| Fecha de publicación: | 2013 |
| País: | España |
| Institución: | Consejo Superior de Investigaciones Científicas (CSIC) |
| Repositorio: | DIGITAL.CSIC. Repositorio Institucional del CSIC |
| OAI Identifier: | oai:digital.csic.es:10261/415148 |
| Acceso en línea: | http://hdl.handle.net/10261/415148 https://api.elsevier.com/content/abstract/scopus_id/84879894065 |
| Access Level: | acceso abierto |
| Palabra clave: | Rainfall Atmospheric circulation Downscaling Non-stationarity |
| Sumario: | [EN] A model for generating daily spatial correlated rainfall fields suitable for evaluating the impacts of climate change on water resources is presented. The model, termed Stochastic Rainfall Generating Process, is designed to incorporate two major nonstationarities: changes in the frequencies of different precipitation generating mechanisms (frontal and convective), and spatial nonstationarities caused by interactions of mesoscale atmospheric patterns with topography (orographic effects). These nonstationarities are approximated as discrete sets of the time-stationary Stochastic Rainfall Generating Process, each of which represents the different spatial patterns of rainfall (including its variation with topography) associated with different atmospheric circulation patterns and times of the year (seasons). Each discrete Stochastic Rainfall Generating Process generates daily correlated rainfall fields as the product of two random fields. First, the amount of rainfall is generated by a transformed Gaussian process applying sequential Gaussian simulation. Second, the delimitation of rain and no-rain areas (intermittence process) is defined by a binary random function simulated by sequential indicator simulations. To explore its applicability, the model is tested in the Upper Guadiana Basin in Spain. The result suggests that the model provides accurate reproduction of the major spatiotemporal features of rainfall needed for hydrological modeling and water resource evaluations. The results were significantly improved by incorporating spatial drift related to orographic precipitation into the model. © 2013 International Association for Mathematical Geosciences. |
|---|