Premium
Methodological application of quantile mapping to generate precipitation data over Northwest Himalaya
Author(s) -
Devi Usha,
Shekhar Manorama S.,
Singh Gyan P.,
Rao Nalamasu N.,
Bhatt Uma S.
Publication year - 2019
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.6008
Subject(s) - precipitation , standard deviation , quantile , environmental science , terrain , climatology , range (aeronautics) , mean squared error , data set , downscaling , statistics , meteorology , mathematics , geography , geology , cartography , materials science , composite material
Continuous high quality data are critical for weather and climate investigations. Numerous data gaps exist particularly over mountainous regions which limits the ability to construct climatologies and perform trend analysis. This study addresses the issue of sparse precipitation data over Northwest Himalaya (NWH) and fills data voids by applying the quantile mapping (QM) method. QM is applied to observed winter precipitation for a period of 25 years (1991–1992 to 2015–2016) to construct a continuous reliable data set. The first 20 years (1991–1992 to 2010–2011) are used for training and the remaining 5 years (2011–2012 to 2015–2016) are used for validation. In total, 10 stations are available for this study and each one is considered serially as a reference to generate daily precipitation values at the other stations. The mean precipitation of NWH region is constructed by considering the mean of all the stations. Standard statistical measures like root mean square errors, standard deviation, skill score and its decompositions are applied to evaluate the generated datasets. Based on statistical analysis, the Kanzalwan station, located in Great Himalaya range, is one of the best performing reference stations for generating precipitation values over NWH. The statistical measures of this station show the highest skill scores, lowest root mean square error and lowest standard mean errors for all winter months except January. This study provides a successful application of QM to generate precipitation data for climate analysis over the complex terrain of the Himalaya region.