Comparison of Areal Snow Storage Sampling Procedures for Rangeland Watersheds

Abstract

The accuracy of photogrammetry in determining snow depth in mountainous rangeland watersheds was evaluated on a 0.41-km2 subbasin of the Reynolds Creek Experimental Watershed, located in the Owyhee Mountains of southwestern Idaho. Random checking of over 50 points indicated that at a photo scale of 1:6000, snow depths were determined with a standard error of ± 15 cm for a mean snow depth of 1.2 m. On the average, only 6% of the snow depths less than 15.2 cm were photogrammetrically determined to be negative, and these were generally during the late melt season. The lag time between photography and the usable result and the need for a field survey to set ground control for each flight relegates this technique to a research tool, rather than an operational forecasting tool. Preliminary evaluation of snow water content on the watershed showed the water content varied according to aspect and deep drift locations. The deep drifts usually had a 6% greater snow water content than the nondrift areas. Simple random, random stratified and two systems of square grids orientated in different directions were tested to determine the best sampling system to determine mean areal snow depth for a watershed. The grid system orientated in the direction of the predominant wind required fewer samples to produce the same accuracy for the snow cover ranging from 100 to 17%.

Mapping snow depth from manned aircraft on landscape scales at centimeter resolution using structure-from-motion photogrammetry

“…A technique that has received considerable attention in recent years is to measure the elevation of the snow surface by airborne or groundbased lidar and subtract from this the snow-free surface elevation, with the difference interpreted as snow depth (Deems et al, 2013;Fassnacht and Deems, 2006;Hopkinson et al, 2004;Prokop, 2008). Operating on the similar principles of repeat or overlapping coverage, but pre-dating lidar studies by 30 years, photogrammetry has also been used to produce snow depth maps (Cline, 1994;König and Sturm, 1998;Lee et al, 2008;McKay, 1968;Najibi and Arabsheibani, 2013;Otake, 1980;Rawls et al, 1980;Yan and Cheng, 2008), including using stereo-imagery from opto-electronic linescanners incorporating near-IR wavelengths in addition to RGB (Bühler et al, 2014;Buhler et al, 2015).…”

Mapping snow-depth from manned-aircraft on landscape scales at centimeter resolution using Structure-from-Motion photogrammetry

“…A technique that has received considerable attention in recent years is to measure the elevation of the snow surface by airborne or groundbased lidar and subtract from this the snow-free surface elevation, with the difference interpreted as snow depth (Deems et al, 2013;Fassnacht and Deems, 2006;Hopkinson et al, 2004;Prokop, 2008). Operating on the similar principles of repeat or overlapping coverage, but pre-dating lidar studies by 30 years, photogrammetry has also been used to produce snow depth maps (Cline, 1994;König and Sturm, 1998;Lee et al, 2008;McKay, 1968;Najibi and Arabsheibani, 2013;Otake, 1980;Rawls et al, 1980;Yan and Cheng, 2008), including using stereo-imagery from opto-electronic linescanners incorporating near-IR wavelengths in addition to RGB (Bühler et al, 2014;Buhler et al, 2015).…”

Precipitation

“…For small-scale roughness elements, random or stratified point sampling (e.g., snow courses) in the area provides an accurate estimation of the snow depth. For large scale roughness elements that typify rangeland and forest watersheds, point sampling on snow courses may be unreliable for representing the total water equivalent of the areal snowpack (Rawls et al, 1980), although the snow course could be suitable for use as an index of runoff.…”