Adjustment of measurement errors to reconcile precipitation distribution in the high‐altitude Indus basin

Abstract

Precipitation in the high-altitude Indus basin governs its renewable water resources affecting water, energy and food securities. However, reliable estimates of precipitation climatology and associated hydrological implications are seriously constrained by the quality of observed data. As such, quantitative and spatiotemporal distributions of precipitation estimated by previous studies in the study area are highly contrasting and uncertain. Generally, scarcity and biased distribution of observed data at the higher altitudes and measurement errors in precipitation observations are the primary causes of such uncertainties. In this study, we integrated precipitation data of 307 observatories with the net snow accumulations estimated through mass balance studies at 21 major glacier zones. Precipitation observations are adjusted for measurement errors using the guidelines and standard methods developed under the WMO's international precipitation measurement intercomparisons, while net snow accumulations are adjusted for ablation losses using standard ablation gradients. The results showed more significant increases in precipitation of individual stations located at higher altitudes during winter months, which are consistent with previous studies. Spatial interpolation of unadjusted precipitation observations and net snow accumulations at monthly scale indicated significant improvements in the quantitative and spatio-temporal distribution of precipitation over the unadjusted case and previous studies. Adjustment of river flows revealed only a marginal contribution of net glacier mass balance to river flows. The adjusted precipitation estimates are more consistent with the corresponding adjusted river flows. The study recognized that the higher river flows than the corresponding precipitation estimates by the previous studies are mainly due to underestimated precipitation. The results can be useful for water balance studies and bias correction of gridded precipitation products for the study area.

Comparison of modeled snow properties in Afghanistan, Pakistan, and Tajikistan

“…Several climate studies of the Indus have focused on using lower-elevation precipitation gauges, which are then used to spatially interpolate basin-wide precipitation. Dahri et al (2016) and Dahri et al (2018) have assembled perhaps the largest collection of climatological measurements covering the AKAH region, mostly based on gauge measurements, as part of a study on the hydrometeorology of the Indus Basin. Using undercatch corrections based on wind, often from reanalysis, they increased precipitation estimates by 21 % on average throughout the Indus Basin (Dahri et al, 2018).…”

“…Dahri et al (2016) and Dahri et al (2018) have assembled perhaps the largest collection of climatological measurements covering the AKAH region, mostly based on gauge measurements, as part of a study on the hydrometeorology of the Indus Basin. Using undercatch corrections based on wind, often from reanalysis, they increased precipitation estimates by 21 % on average throughout the Indus Basin (Dahri et al, 2018). For example, in the Gilgit sub-basin, they find an unadjusted precipitation estimate of 582 mm yr −1 , adjusted to 787 mm yr −1 , a 35 % increase.…”

Cross-validating precipitation datasets in the Indus River basin

“…Therefore, a comprehensive evaluation of the basin-wide water cycle is needed. Studies that have addressed this issue have stressed the uncertainties inherent in the observed precipitation (Singh et al, 2011;Gardelle et al, 2012;Immerzeel et al, 2015;Wang et al, 2017;Dahri et al, 2018).…”

“…Beside the risk of corruption or missing values in the reporting process, it has been demonstrated that rain gauges can underestimate precipitation (Sevruk, 1984;Goodison et al, 1989). The main source of underestimation is the wind-driven under-catchment that can reach up to 50 % during snowfall (Goodison et al, 1989;Adam and Lettenmaier, 2003;Wolff et al, 2015;Dahri et al, 2018), but also includes the wetting of the instrument, evaporation before measuring, and splashing out (WMO, 2008). Dahri et al (2018) used the guidelines from the World Meteorological Organization (WMO) to re-evaluate the precipitation measured from hundreds of rain gauges in the upper Indus and found the underestimation to be between 1 % and 65 % for each station, and 21 % across the basin.…”

“…Rain gauges are not only scarce in mountainous areas, but their location is also biased. In order to be accessible all year long, they are gener-ally situated at the bottom of valleys, and these locations appear to be significantly drier than locations at altitude (Archer and Fowler, 2004;Ménégoz et al, 2013;Immerzeel et al, 2015;Dahri et al, 2018), which means that the interpolation method underestimates precipitation in the surrounding mountains.…”