WSP04999 - Laserfiche WebLink

Home Browse Search

eswc 14.7 Supp = 1--= 1--=.34 eswn 22.3 which is the same at both points I and 2 because of the constant R.H. assumption. For this situation the model estimates a suppression at the upwind point (applicable to a small reservoir) which is 19 percent low. At the downwind point (more applicable to a large reservoir) the model suppression is 8 percent low. Figure 5 shows a graphic interpretation of error introduced by the model which is more relevanl than the particular magnitudes of error in the scenario selected for the figure. A comparison of the Dalton ratios and the model ratios in Figure 5 are instructive. The ~l/ 6enl ratio shown in the figure can be modified to equal the model ratio by adding eal to both the numerator and the denominator. In other words, the error introduced by the model at a small reservoir is directly proportional to the vapor pressure of the air. The model will always be conservative for. a small reservoir because adding a constant to both the numerator and denominator of a fraction that is less than unity must increase the fraction. On a large reservoir the model increases the numerator by an amount which is Jess than that added to the denominator thereby simulating the Dalton equation more accurately. Although better model accuracy is expected for large reservoirs, error on the conservative side is not assured. One conclusion that is app~rent from Figure 5 is that the model simulates suppression rather well at low humidity but introduces very large error at high humidity. This is not considered to be a serious drawback because the only areas where evaporation suppression is likely to be worthwhile are arid regions. And the time when accuracy of the model is most important is the dry summer period. Energy budget cODllderatloDl The previous discussion of potential suppres- sion has been essentially in context of relationship between parameters at any given point in time. The integration of these effects on a reservoir during a season or a series of years requires the considera. tion of heat addition and loss sources over time. A destratification study by the U.S. Geological Survey on Lake Wohlford. California, included the only previously published attempt to analyze the effect of thermal mixing on evaporation (Koberg et al.. 1965). Even though the reservoir was less than 50 feet deep, a 5 percent net savings in evaporation was computed. The researchers had expected a negative effect in the fall which equaled the suppression effect during the early summer. They explained the apparently unexpected net savings as being due to draw down during the fall. In addition to draw down, however, the prelimi- nary mode' for Utah reservoirs indicates that a relaled but more important factor many be the increased temperature of released water (which may be operative even under negative draw down conditions). The hyPothesis of the authors in this regard (which anticipates a net suppression rather than an annual balance) is two.fold. 1. Thennal mixing achieves a significant increase in temperature of water flowing from the reservoir outlet (assuming the outlet is near the reservoir bottom or at least below the thermocline), and therefore a net decrease in reservoir heat is accomplished by flow from the reservoir. 2. On reservoirs for which winter carryover storage is a minor part of annual storage, residual winter heat is unimportant. In this situation, a May to October suppression of evaporation (the irrigation season) for example, is all that may be of concern because spring runoff will fill the impoundment anyway (and will dominate subse- quent water temperatures). On reservoirs which have high carryover stor- age factors however, the comparison between heat added by suppression and increased heat losl from the outlet will be of key importance in determining the net annual evaporation suppression. This can best be visualized by considering the significant sources of heat flux in a reservoir energy budget. U.S. Geological Survey researchers have defined nine such variables as constituting energy budget parameters on studies of Lake Mead (Harbeck et aI., (958), Lake Colorado City (Harbeck, et al.. 1959), and the Salton Sea (Hughes, 1967). The USGS parameters are shown in Figure 6. In these papers the energy budget was defined as follows: Qs - Or +Oa - Oar - Obs + Ov - Oe - Oh - Ow = Net change (see Figure 6). The first four terms are respectively the incoming and reflected solar radiation and atmospheric long-wave radiation. Although these four are major items in the energy budget, they can be ignored in the proposed model because they are independent of changes in surface temperature (except in the very minor portion of the surface area where the albedo may be changed by air bubbles). Qbs is the long wave radiation emitted by the body of water. This is an important component of the energy budget which is a function of surface 12