Identification of dominant sources of sea level pressure for precipitation forecasting over Wales

0
- By: ,

Courtesy of IWA Publishing

Downscaling methods are utilized to assess the effects of large scale atmospheric circulation on local hydrological variables such as precipitation and runoff. In this paper, a methodology of statistical downscaling using a support vector machine (SVM) approach is presented to simulate and predict the precipitation using general circulation model (GCM) data. Due to the complexity and issues related to finding a relationship between the large scale climatic parameters and local precipitation, the climate variables (predictors) affecting monthly precipitation variations over Wales are identified using a combination of the methods including the principal component analysis (PCA), fuzzy clustering, backward selection, forward selection, and Gamma test (GT). The effectiveness of those tools is illustrated through their implementations in the case study. It has been found that although the GT itself fails to identify the best input variable combination, it provides useful and narrowed-down options for further exploration. The best input variable combination is achieved by the GT and forward selection method. This approach can be a useful way for assessing the impacts of climate variables on precipitation forecasting.

Customer comments

No comments were found for Identification of dominant sources of sea level pressure for precipitation forecasting over Wales. Be the first to comment!