Deriving Bias and Uncertainty in MERRA-2 Snowfall Precipitation Over High Mountain Asia

Yufei Liu,Steven A Margulis

doi:10.3389/feart.2019.00280

Yufei Liu, Steven A Margulis

Open Access

https://doi.org/10.3389/feart.2019.00280

Copy DOI

Abstract

A Bayesian approach to estimate bias and uncertainty in snowfall precipitation from MERRA-2 and other precipitation products was applied over High Mountain Asia (HMA), using a newly developed snow reanalysis method. Starting from an ‘uninformed’ prior probability distribution, a posterior scaling factor applied to MERRA-2 snowfall was derived by constraining model-based estimates of seasonal snow accumulation and ablation over the water year (WY) with fractional snow covered area (fSCA) measurements derived from Landsat and MODIS (MODSCAG). Several sub-domains (nine representative 1° by 1° tiles) across HMA were examined over the period WYs 2001-2015 and compiled into an uncertainty parameterization where a lognormal distribution was fitted to the empirical posterior distribution with a mean of 1.54 (median of 1.19) and coefficient of variation of 0.83, indicating that MERRA-2 underestimates snowfall on average by ~54% with sizeable uncertainty. For reference, the uncertainties in snowfall precipitation from the ERA5 and APHRODITE-2 precipitation produces were also evaluated, and these products were found to underestimate snowfall, on average by a factor around 1.78 and 3.34 (with median scaling factors of 1.42 and 2.51) respectively. The results indicate that snowfall precipitation at high-elevations dominated by snowfall is underestimated in most existing products, especially in the gauge-based APHRODITE-2 product, where the biases were also found to exhibit geographical variations with the largest underestimation in monsoon-influenced high-elevation tiles. The derived MERRA-2 uncertainty model is being used to develop a full domain-wide HMA snow reanalysis, which will shed further light onto the space-time variations in snowfall biases in these products.

Highlights

Accurate snowfall information is vital for hydrological modeling in snow-dominated regions, as it directly affects the estimation of snow water equivalent (SWE), and influences streamflow prediction fed by snowmelt
Based on that characterization and inter-comparison with other products (e.g., APHRODITE-2 and ERA5) we address the following questions: Is MERRA-2 snowfall biased over High Mountain Asia (HMA) and how can its bias and uncertainty be parameterized? Is snowfall biased in other gridded precipitation products and to what extent? How do the snowfall biases vary spatially for these products and what is the spatial-temporal distribution of snowfall?
The climatology maps of the interpolated MERRA-2 precipitation, the downscaled MERRA2 snowfall, the posterior snowfall from the snow reanalysis and the posterior b from the snow reanalysis at each of the test tiles are shown in Figure 4

Summary

Introduction

Accurate snowfall information is vital for hydrological modeling in snow-dominated regions, as it directly affects the estimation of snow water equivalent (SWE), and influences streamflow prediction fed by snowmelt. Meteorological stations are sparsely located in this region, generally located at lower elevations, and data from these stations can lack representativeness of precipitation at unmonitored locations, most notably higher elevations (Winiger et al, 2005; Palazzi et al, 2013) Satellite observations such as TRMM (Tropical Rainfall Measuring Mission; Huffman et al, 2007) can provide spatially continuous precipitation estimates, but have been found to be uncertain and potentially biased in the HMA region (Anders et al, 2006; Bookhagen and Burbank, 2006; Andermann et al, 2011; Hussain et al, 2017; Khan and Koch, 2018; Khan et al, 2018). Both remote sensing and in situ data sources generally perform worse at characterizing snowfall compared to rainfall, often with significant underestimates of snowfall rates, because their instruments are mainly designed to measure liquid rainfall rather than snow or ice (Anders et al, 2006; Viste and Sorteberg, 2015)

Objectives

Methods

Results

Conclusion