Evaluating the Quality of “Citizen Scientist” Snow Albedo Observations for use in Climate System Studies

The purpose of this study is to evaluate the quality of snow albedo estimates derived from observations reported by volunteer observers for use in climate system studies. Networks of volunteer observers have provided valuable, high-density weather observations for decades and continue to provide daily snowfall and snow depth measurements at thousands of locations across the US. Although snow-surface reflectively is an important variable in winter climate processes, in situ snow albedo data are collected at far fewer locations and existing sites are widely scattered. To address this issue, a pilot network of 16 volunteers recruited from the New Hampshire network of Community Collaborative Rain Hail and Snow (CoCoRaHS) observers was established for the 2011/12 winter season to collect detailed snow-pack observations from which snow-surface albedo was estimated. Observers were trained and provided with an instrument kit at a cost of approximately $700 per observer that included (1) measuring sticks for depth of new snow and snow on the ground, (2) a set of aluminum snow tubes and CCi HS digital hanging scale for measuring the weight of snow on the ground, and (3) a mounted Apogee MP-200 pyranometer for shortwave radiation measurements used to estimate albedo. Volunteers made daily observations of general weather conditions (sky cover), snow depth and weight, and incoming and reflected shortwave radiation within ± 1 hour of solar noon from November 23, 2011 through March 15, 2012. An initial evaluation of albedo estimates derived from volunteer observations showed that the New Hampshire albedo estimates were within the range of values reported for similar mid-latitude locations. Analyses showed that albedo estimates derived from the Apogee MP-200 measurements were strongly correlated (r² > 0.99) with albedo measurements obtained using a standing Kipp and Zonen CMA6 albedometer. Evaluation of the MP-200 measurements against the CMA6 showed a net bias of +0.05, which is within the range of absolute accuracy for climate system modeling, and this bias was reduced to +0.01 when calibrated with the CMA6. Comparisons with an ASD FieldSpec4 spectrometer measurements indicated that the relationship between observed albedo and snow depth/density were in good agreement with the spectral albedo measurements and reasonably represent the physical characteristics of the snow pack at each station location. Given the acceptable accuracy of the low-cost MP-200 and the high skill of the volunteers, this type of volunteer observer network has the potential to provide research quality snow albedo data for use in climate system studies. Therefore, this network will be expanded to include 25 volunteer observation sites across New Hampshire for the 2012/13 winter season.