Fine-scale rainfall data is important for many hydrological applications including soil moisture modelling, flood prediction and land-slide risk assessment. However, direct approaches to obtain such data are generally difficult, usually because the required sensors are expensive and a fine grained network is needed. Evidently, this is unfeasible in many areas of the globe. In recent times, satellite based measuring methods offer a new opportunity to gain unprecedented coverage, however, at the cost of a decreased resolution. This has given rise to downscaling methods, a group of statistical and dynamical methods that allow for an increase in resolution of coarse-scale imagery.
In recent work, a copula based approach to downscaling has been explored. In this work, it was shown that a copula-based framework can approximate the sub-pixel distribution of coarse-scale pixels. However, the methodology of the study was a proof-of-concept and lacked practical applicability. Here, we present work where we expand upon the previous work by fitting a parametrical copula and significantly expanding the dataset. This allows us to examine the behaviour of the dependence within and between storms of different types (stratiform, convective). Moreover, it is possible to examine the scaling behaviour of the dependence and extrapolate it to scales not measured. Finally, the practicality of the framework is improved to allow for practical application of the framework within other downscaling methods.
1. van den Berg, M. J., Vandenberghe, S., De Baets, B., and Verhoest, N. E. C.: Copula-based downscaling of spatial rainfall: a proof of concept, Hydrol. Earth Syst. Sci. Discuss., 8, 207-241, 2011.
2. Vandenberghe, S., N. E. C. Verhoest, and B. De Baets (2010), Fitting bivariate copulas to the dependence structure between storm characteristics: A detailed analysis based on 105 year 10 min rainfall, Water Resour. Res., 46, W01512
Keywords: Copula; Downscaling; Rainfall
Biography: The author has a passion for statistical techniques and their practical applications. This is focussed on the augmentation of already existing data, to further knowledge of Hydrology and Earth surface processes.