Statistical downscaling of general circulation model outputs to precipitation, evaporation and temperature using a key station approach
- Authors: Sachindra, Dhanapala , Huang, Fuchun , Barton, Andrew , Perera, Bimalka
- Date: 2016
- Type: Text , Journal article
- Relation: Journal of Water and Climate Change Vol. 7, no. 4 (2016), p. 683-707
- Full Text: false
- Reviewed:
- Description: Using a key station approach, statistical downscaling of monthly general circulation model outputs to monthly precipitation, evaporation, minimum temperature and maximum temperature at 17 observation stations located in Victoria, Australia was performed. Using the observations of each predictand, over the period 1950-2010, correlations among all stations were computed. For each predictand, the station which showed the highest number of correlations above 0.80 with other stations was selected as the first key station. The stations that were highly correlated with that key station were considered as the member stations of the first cluster. By employing this same procedure on the remaining stations, the next key station was found. This procedure was performed until all stations were segregated into clusters. Thereafter, using the observations of each predictand, regression equations (inter-station regression relationships) were developed between the key stations and the member stations for each calendar month. The downscaling models at the key stations were developed using reanalysis data as inputs to them. The outputs of HadCM3 pertaining to A2 emission scenario were introduced to these downscaling models to produce projections of the predictands over the period 2000-2099. Then the outputs of these downscaling models were introduced to the inter-station regression relationships to produce projections of predictands at all member stations.
Least square support vector and multi-linear regression for statistically downscaling general circulation model outputs to catchment streamflows
- Authors: Sachindra, D. A. , Huang, Fuchun , Barton, Andrew , Perera, Bimalka
- Date: 2013
- Type: Text , Journal article
- Relation: International Journal of Climatology Vol. 33, no. 5 (2013), p. 1087-1106
- Full Text: false
- Description: This study employed least square support vector machine regression (LS-SVM-R) and multi-linear regression (MLR) for statistically downscaling monthly general circulation model (GCM) outputs directly to monthly catchment streamflows. The scope of the study was limited to calibration and validation of the downscaling models. The methodology was demonstrated by its application to a streamflow site in the Grampian water supply system in northwestern Victoria, Australia. Probable predictors for the study were selected from the National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data set based on the past literature and hydrology. Probable variables that displayed the best significant correlations, consistently with the streamflows over the entire period of the study (1950-2010) and under three 20-year time slices (1950-1969, 1970-1989 and 1990-2010) were selected as potential predictors. To better capture seasonal variations of streamflows, downscaling models were developed for each calendar month. The standardized potential predictors were introduced to the LS-SVM-R and MLR models, starting from the best correlated three and then, others one by one, based on their correlations with the streamflows, until the model performance in validation was maximized. This stepwise model development enabled the identification of the optimum number of potential variables for each month. The model calibration was performed over the period 1950-1989 and validation was done for 1990-2010. LS-SVM-R model parameter optimization was achieved using simplex algorithm and leave-one-out cross-validation. The MLR models were optimized by minimizing the sum of squared errors. In both modelling techniques, validation was performed as an independent simulation. In calibration, LS-SVM-R and MLR models displayed equally good performances with a trend of under-predicting high flows. During validation, LS-SVM-R outperformed MLR, though both techniques over-predicted most of the streamflows. It was concluded that LS-SVM-R is a better technique for statistically downscaling GCM outputs to streamflows than MLR, but still MLR is a potential technique for the same task. Copyright © 2012 Royal Meteorological Society.