Laserfiche WebLink
RGDSS Memorandum <br />Final <br />To: Ray Bennett, Ray Alvarado, Andy Moore <br />From: LRCWE, Erin Wilson and Janet Williams <br />Subject: Rio Grande Historic Crop Consumptive Use -Climate Data <br />Date: October 28, 1999 <br />Overview <br />This memorandum presents the approach and results obtained for the Consumptive Use and <br />Water Budget Component Subtask 3.1 -Select and Fill Key Climate Stations. <br />The estimation of potential crop consumptive use requires climate data for temperature, <br />precipitation and frost dates. For the Rio Grande Historic Crop Consumptive Use Analysis and <br />the Rio Grande Water Resource Planning Model Application, climate data sets were generated <br />through a data centered approach. <br />General Approach <br />For the generation of climate data sets for RGDSS modeling, linear regression was used to fill <br />the temperature data sets, and long-term averages were used to fill missing precipitation and frost <br />date data for calendar years 1950 through 1997. The filling processes are described in more <br />detail below. <br />Filling of Missing Values in the Temperature Data Sets <br />Linear regression requires dependant (to be filled) and independent (to be used as the basis for <br />filling) data sets. Two weather stations in the Rio Grande Basin were selected for use as primary <br />independent data sets based on their completeness of historic record during the 1950 through <br />1997 period and their goodness of fit coefficients (r~) in regressions with the remaining stations. <br />The two stations selected for use as potential primary independent data sets were: <br />• Alamosa WSO AP (NWS No. 130, 1948-1997) with missing values for 12 months during the <br />1950-1997 period. <br />• Del Norte (NWS No. 2184, 1948-1997) with missing values for 43 months during the 1950- <br />1997 period. <br />Missing data in the Alamosa temperature data set were filled using Del Norte as the independent <br />data set, based on its high correlation coefficient with Alamosa and because there were no <br />missing monthly values between the two stations. The regression between Alamosa and Del <br />Norte was performed using the DMI tstool. Regressions were then performed between unfilled <br />temperature series for Alamosa and other stations, and between filled temperature series for <br />appendD_cropcu.doc D-1 of D-9 10/28/99 <br />