Presenting Snow Grain Size and Shape Distributions in Northern Canada Using a New Photographic Device Allowing 2D and 3D Representation of Snow Grains

Alexandre Langlois,Alexandre Roy,Alain Royer,Martin Durocher,Benoit Montpetit

doi:10.3389/feart.2019.00347

Abstract

Geophysical properties of snow are known to be sensitive to climate variability and are of primary importance for hydrological and climatological process simulations. Numerous studies using passive microwaves have attempted to quantify snow from space, but the methods suffer from poor spatial resolution retrievals, combined with a great sensitivity to snow grain morphology. Those issues motivated work using active microwaves that are now core to space mission concept proposals currently under development. However, a clear limitation remains with regards to snow microstructure contribution to backscattering, especially in large depth hoar layers typical of polar snowpacks. This leads to difficulties retrieving Snow Water Equivalent (SWE) from space or developing radiative transfer models owing to a lack of field observations of snow microstructure. This paper presents an innovative technique to measure various snow grain metrics in the field where micro-photographs of snow grains are taken under angular directional LED lighting. The projected shadows are digitized so that a 3D reconstruction of the snow grains is possible and distribution functions can be proposed for various snow grain metrics and grain types. This device, dLED, has been used in several field campaigns and a very large dataset was collected and is presented in this paper. Distribution histograms from >160,000 digitized grains were produced for each metric for all grains considered as a whole dataset (unclassified), and also for each grain type: 1) Defragmented/broken (DF), 2) Depth Hoar (DH), 3) Facets (F), 4) Rounds (R) and 5) Precipitation Particles (PP). We selected distribution functions for each metric per grain time by analysing L-moment diagrams that summarize the shape of a probability distribution. Our results show that the logarithmic Kappa (LKAP) distribution is well suited to explain the snow grain metric distribution for each grain type. Location, scale and shape parameter values for each distribution are presented and a comparison with values derived from our shortwave infrared laser device, the InfraRed Integrating Sphere (IRIS), is provided. A discussion is presented on the pros and cons of the dLED and the use of the distributions presented in this paper for microwave radiative transfer modeling work.

Highlights

In the context of global climate change observed over the past four decades in northern regions, numerous studies have focused on the retrieval of surface state variables to monitor the rate and amplitude of observed changes (Takala et al, 2011; Brown and Derksen, 2013; Estilow et al, 2015)
The underestimation remains, but results suggest that for lower surface specific area (SSA) snow grains, the dLED could retrieve SSA with reasonable accuracy
The dataset was collected over a period of 6 months using a new device allowing the digitization of projected snow grain shadows under LED illumination

Summary

Introduction

In the context of global climate change observed over the past four decades in northern regions, numerous studies have focused on the retrieval of surface state variables to monitor the rate and amplitude of observed changes (Takala et al, 2011; Brown and Derksen, 2013; Estilow et al, 2015). The Arctic is warming at more than twice the rate of lower latitudes, leading to a decrease in sea ice cover (Serreze and Stroeve, 2015), glacier mass balance (Papasodoro et al, 2015), permafrost extent (Schuur et al, 2015) and snow cover (Derksen and Brown, 2012) This is of particular relevance in a context where snow covers up to 40 million km during winter in North America and supports freshwater supplies for consumption, agriculture and hydroelectricity. Related uncertainties from space retrievals and the development of complex thermodynamic multilayer snow and microwave emission models motivated several studies on the development of new approaches to quantify snow grain (e.g., Domine et al, 2006; Matzl and Schneebeli, 2006; Langlois et al, 2010; Montpetit et al, 2012) given the lack of field measurements

Objectives

Results

Discussion

Conclusion