Measurement of a Changing Moisture Climate

John Hallett Desert Research Institute,Reno NV. Understanding processes leading to a changing climate may be characterized in terms of selected variables(as precipitation) whose very measurment may lead to signifcant controversy. Greater insight and objectivity is obtainable by comparing probability functions(how much precipitation falls at what rate) derived from space and/or time selected on physical grounds, as a frontal passage or a whole season.This approach requires new instrument development to provide a high resolution in space/time consistent with the Poisson uncertainty of collection of individual precipitation particles and the patience to wait for a sufficient time from minutes to decades.Design criteria necessarily involve collocated instruments(as Hooke realized many years ago) and high resolution (in time and space)to be only degraded by smoothing, but never improved. The challenge is accepted to design and deploy a new instrument, the hotplate,to provide both the vertical flux of precipitation (rain, snow or a mix thereof) and also the horizontal wind speed;the system is to operate independently of wind direction. The design led to a 13cm diameter plate maintained at constant temperature, with power requirements giving precipitation rate (through rate of latent heat evaporation) and wind speed, overlaying and shielding an identical parallel reference plate 3cm underneath to provide the wind speed alone. The plates operate at 100C (having an added advantage of discouraging birds)and a time constant of ten seconds. The system may be duplicated in the vertical, as at a height of 1m and 2m to provide a measure of blowing snow and Richardson number as a link to stability and turbulence. The systems have worked well for a number of years both under winter and tropical storm conditions, for precipitation from a fraction mm per hour to > 60mm per hour.The overall advantage of the approach lies in the comparison of the observered probablity relationships with prediction of probability distributions as a test of related models similarly constrained. (As part of a collaborative project with German Vidaurre, Texas A&M, Robert Black, NOAA AOML Miami FL, Roy Rasmussen, NCAR CO,Rick Purcell,DRI Reno NV, Dave Simeral/Greg McCurdy, DRI Reno NV).