Intercomparison of Recording and Standard Nonrecording U.S. Gauges

Abstract

Existing adjustment procedures for precipitation measured by the 8-in. standard nonrecording rain gauge (SNRG) take into account gauge undercatch mostly due to wind-induced turbulence over its orifice. The goal of this paper is to ?reduce? the measurements of the two most commonly used U.S. recording rain gauges (SRGs) to those for a hypothetical SNRG at the same site and use established relationships to adjust their readings to less biased values. This will allow gauge-specific adjustments for each gauge/precipitation type to secure less biased homogeneous precipitation time series and facilitate blending of precipitation information from different sources/networks. A significant portion of the U.S. rain gauges are automated recording gauges. A majority of the gauges of the National Oceanic and Atmospheric Administration Hourly Precipitation Network are unshielded Fischer and Porter recording rain gauges (FP) that have gradually replaced weighing recording rain gauges (WG). There are more than 1400 sites where the unshielded SNRG and FP are collocated (and more than 100 sites for SNRG and WG). Daily precipitation from these sites from 1982 to 1996 were partitioned into three categories: rainfall, frozen, and mixed precipitation. The hourly records from recording gauges were totaled over the 24-h-long period specific to each station to match the observation times of the SRGs and compared with the daily SNRG records. The comparison of the rain and snow catch of WG and SNRG shows nearly equivalent measurements. Mixed precipitation is the most troublesome type to register for WG and averages 7% less than collocated SNRG. Rain catch of the FP averages 5% less than SNRG and is not strongly related to climate conditions and/or surroundings of meteorological sites. Also, FP observations of frozen precipitation can be as high as 105% and as low as 70% compared to SNRG with an average of 90% (85% over the Great Plains). In open, windy locations, the SNRG can underestimate frozen precipitation by 50% or more. The authors? estimates indicate that FP underestimates frozen precipitation even more than SNRG, and the level of the underestimation is a function of wind speed and gauge exposure.