SPDSS_Task59-1_Develop Locally Calibrated Blaney-Criddle Crop Coefficients

Home Browse Search

It is important when evaluating data from different lysimeters to recognize differences between haying practices, irrigated environment around the lysimeters, means of filling the lysimeter, changes of vegetative matter, operational problems, etc. Examples of inconsistencies found include: The Division VI lysimeters are filled manually rather than automatically filling from a reservoir controlled by a float. According to Division VI staff, the sites were flooded at least once a month or when necessary to prevent drought stress to the plants (an attempt is made to irrigate before wilting occurs). These lysimeters are also surrounded by large non-irrigated pasture that may make them unrepresentative of irrigated meadows. While one would typically believe that the location of the lysimeter sites in non-irrigated meadows would increase the measured consumptive use, we believe that the manual filling of the lysimeters based on visual observation produces measured consumptive use more similar to an actual consumptive use value (with limited irrigation) rather than potential consumptive use. The South Park lysimeter data used by LRCWE (1993) was not adjusted to reflect haying practices. Adjustment for haying (which usually occurs in early Augustin South Park) would reduce the averaged coefficients for August and September. The derivation of the crop coefficients also appears to have included less detailed review for consistency and adjustment when compared to the Walter 1990 study. While the Walter Study considered some of the same lysimeter data, Walter discarded from further consideration a number of the sites that were not considered representative of high altitude irrigated meadows and adjusted the crop coefficients of other sites to be consistent with a hayed meadow. • The Gunnison -Kruse crop coefficient pattern has a different shape than most of the crop coefficient curves derived from high altitude lysimeter studies which tend to peak in June, bottom out in August (with harvest) and then recover some after harvest in September. The Kruse study did not provide an explanation for the Gunnison bowl-shaped coefficient curve. Both the Walter study for the South Park area and the Smith study for the Upper Gunnison River Basin appear to have made a reasonable effort to identify and exclude lysimeter data that was not reflective of irrigated meadows, and adjust data for consistency (i.e. haying practices). The Walter and Smith crop coefficient curves shown in Figure 1 each represent an average of lysimeter results from several different sites. A summary of the individual lysimeter studies reviewed by Walter are shown below in Figure 2 and the individual site results from the Smith study are shown in Figure 3. So while the Walter and Smith curves shown in Figure 1 appear to be at the upper extent of the studies reviewed, they are actually the average of a subset of studies and individual site curves. These are the two most comprehensive and complete lysimeter studies reviewed and while producing independent derivations of crop coefficients, the resulting average monthly crop coefficients are very similar. Application of either set of coefficients with the original Blaney-Griddle method would be reasonable for SPDSS, however, due to the level of documentation provided with the Walter study we recommend the South Park -Walter coefficients be used for SPDSS (Table 1 shown in bold). The South Park -Walter study showed Task59_l.doc (updated January 11, 2008) 4 of 32