Combination of remote sensing data products to derive spatial climatologies of “degree days” and downscale meteorological reanalyses: application to the Upper Indus Basin

Abstract:

Lack of observations for the full range of required variables is a critical reason why many cryosphere-dominated hydrological modelling studies adopt a temperature index (degree day) approach to meltwater simulation rather than resolving the full surface energy balance. Thus spatial observations of "degree days" would be extremely useful in constraining model parameterisations. Even for models implementing a full energy balance, "degree day" observations provide a characterisation of the spatial distribution of climate inputs to the cryosphere-hydrological system.

This study derives "degree days" for the Upper Indus Basin by merging remote sensing data products: snow cover duration (SCD), from MOD10A1 and land surface temperature (LST), from MOD11A1 and MYD11A1. Pixel-wise “degree days” are calculated, at imagery-dependent spatial resolution, by multiplying SCD by (above-freezing) daily LST. This is coherent with the snowpack-energy-to-runoff conversion used in temperature index algorithms. This allows assessment of the spatial variability of mass inputs (accumulated snowpack) because in nival regime areas – where complete ablation is regularly achieved – mass is the limiting constraint. The GLIMS Randolph Glacier Inventory is used to compare annual totals and seasonal timings of “degree days” over glaciated and nival zones. Terrain-classified statistics (by elevation and aspect) for the MODIS “degree-day” hybrid product are calculated to characterise of spatial precipitation distribution.

While MODIS data products provide detailed spatial resolution relative to tributary catchment areas, the limited instrument record length is inadequate for assessing climatic trends and greatly limits use for hydrological model calibration and validation. While multi-decadal MODIS equivalent data products may be developed in the coming years, at present alternative methods are required for “degree day” trend analysis. This study thus investigates the use of the hybrid MODIS “degree day” product to downscale an ensemble of modern global meteorological reanalyses including ERA-Interim, NCEP CFSR, NASA MERRA and JRA-55 which overlap MODIS instrument record. This downscaling feasibility assessment is a prerequisite to applying the method to regional climate projections.