Statistical downscaling of rainfall: a non-stationary and multi-resolution approach

Rashid, Md. Mamunur and Beecham, Simon and Chowdhury, Rezaul Kabir (2016) Statistical downscaling of rainfall: a non-stationary and multi-resolution approach. Theoretical and Applied Climatology, 124 (3-4). pp. 919-933. ISSN 0177-798X

Abstract

A novel downscaling technique is proposed in this study whereby the original rainfall and reanalysis variables are first decomposed by wavelet transforms and rainfall is modelled using the semi-parametric additive model formulation of Generalized Additive Model in Location, Scale and Shape (GAMLSS). The flexibility of the GAMLSS model makes it feasible as a framework for non-stationary modelling. Decomposition of a rainfall series into different components is useful to separate the scale-dependent properties of the rainfall as this varies both temporally and spatially. The study was conducted at the Onkaparinga river catchment in South Australia. The model was calibrated over the period 1960 to 1990 and validated over the period 1991 to 2010. The model reproduced the monthly variability and statistics of the observed rainfall well with Nash-Sutcliffe efficiency (NSE) values of 0.66 and 0.65 for the calibration and validation periods, respectively. It also reproduced well the seasonal rainfall over the calibration (NSE = 0.37) and validation (NSE = 0.69) periods for all seasons. The proposed model was better than the tradition modelling approach (application of GAMLSS to the original rainfall series without decomposition) at reproducing the time-frequency properties of the observed rainfall, and yet it still preserved the statistics produced by the traditional modelling approach. When downscaling models were developed with general circulation model (GCM) historical output datasets, the proposed wavelet-based downscaling model outperformed the traditional downscaling model in terms of reproducing monthly rainfall for both the calibration and validation periods.


Statistics for USQ ePrint 34204
Statistics for this ePrint Item
Item Type: Article (Commonwealth Reporting Category C)
Refereed: Yes
Item Status: Live Archive
Additional Information: Restricted access to published version in accordance with the copyright policy of the publisher.
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Civil Engineering and Surveying
Date Deposited: 05 Jun 2018 00:21
Last Modified: 23 Aug 2018 03:59
Uncontrolled Keywords: monthly rainfall, continuous wavelet transform, validation period, statistical downscaling, original series
Fields of Research : 04 Earth Sciences > 0401 Atmospheric Sciences > 040105 Climatology (excl.Climate Change Processes)
09 Engineering > 0905 Civil Engineering > 090509 Water Resources Engineering
Socio-Economic Objective: D Environment > 96 Environment > 9603 Climate and Climate Change > 960303 Climate Change Models
Identification Number or DOI: 10.1007/s00704-015-1465-3
URI: http://eprints.usq.edu.au/id/eprint/34204

Actions (login required)

View Item Archive Repository Staff Only