|Johnson, J - USACE/CRREL|
Submitted to: Trans American Geophysical Union
Publication Type: Abstract Only
Publication Acceptance Date: November 18, 2003
Publication Date: November 20, 2003
Citation: Johnson, J. Marks, D., 2003. Methods to identify and correct snow water equivalent pressure sensor measurements errors to improve estimates of water stored as snow. EOS Transactions, v. 84(46) supplement, American Geophysical Union, CD-ROM abstract. Technical Abstract: Snow water equivalent (SWE) pressure sensor measurement errors significantly reduce the ability of water resource managers to accurately predict the amount of available SWE or snowmelt water runoff. SWE errors may occur when the snow/soil interface is at the melting temperature and the differential snowmelt rate between a SWE sensor and the soil exceeds the snow creep rate. A cold, dry snow cover overlying frozen soil will not exhibit SWE sensor errors because the snow and soil temperatures are both below the melting temperature. A delayed SWE sensor response may occur when the SWE accumulation rate exceeds the snow creep rate. Cold snow covers with many ice layers will have a relatively low creep rate while isothermal wet snow will have a relatively high creep rate. This is why a SWE sensor that exhibits a large error just before the onset of wet isothermal conditions will suddenly start to accurately measure SWE. By understanding the physical mechanisms that affect SWE sensor performance the occurrence of SWE errors can be identified and a variety of approaches to correcting them are possible. When only SWE sensor data are available, the occurrence of SWE errors can be identified by excursions in SWE, which can be corrected by interpolating from data on either side of the excursion. A delayed SWE sensor response to a snowstorm is related to the snow creep rate and can be corrected by shifting the SWE record to ensure that the start of the SWE increase and the storm agree. The SWE accumulation rate midway between the beginning and end of the storm should provide a good approximation of the storm's accumulation rate. When additional information on precipitation, snow depth, and snow core SWE are available then relatively simple comparison and interpolation methods can be used to correct errors. Ground temperature or heat flux data can be used to predict SWE sensor performance and correct SWE errors. Air temperature data and a heat flux model of snow and soil may be used to estimate data needed to correct SWE errors.