Scaling characteristics of precipitation data in conjunction with wavelet analysis

Abstract

Summary A large set of monthly precipitation data from 43 stations throughout Texas was employed to investigate the spatial variability in the multiscaling properties of wet and dry spells. Special emphasis was given to dry spells which are related to meteorological droughts. Scaling properties deduced from the analysis of dry spells can be used in drought modeling and multiscale temporal variability of droughts. Using moment scaling exponents, scaling properties of wet and dry spells were examined for a median truncation level. No coherent regional differences were found from the spatial depiction of scaling parameters. Wet and dry spells showed different tendencies in simple scaling and multiscaling throughout the study area. Also, significant low frequency patterns of precipitation were found when the wavelet transform was used. Investigation of the relationship between scaling properties and significant cycles of precipitation data showed that annual cycles may contribute to the occurrence of simple scaling mechanism in wet and dry spell sequences. Such characterization of sequences of wet and dry spells is essential for addressing a multitude of hydrological problems, including estimation of flood and drought frequencies, construction of rainfall and runoff relationships, agricultural planning, to name but a few.