MONTHLY SHALLOW POND EVAPORATION IN IDAHO This Web publication was adapted from
Molnau,Myron, Kpordze, Kojo C.S., and Craine,Katherine L., 1992. Monthly shallow pond evaporation in Idaho. ASAE paper PNW 92-111
ABSTRACT
The design of ponds and lagoons require a water balance for the pond area. One of the items of greatest uncertainty in determining this balance is a measure of the evaporation from the water surface. The monthly evaporation is required by various permitting agencies but the areal distribution of pan evaporation stations is not adequate to allow monthly evaporation amounts to be determined. This paper presents a method whereby the monthly pond evaporation can be calculated using an evaporation map from the NOAA evaporation atlas and monthly percentages of annual evaporation developed from monthly evapotransipration and pan evaporation data.
EXAMPLE OF USE This example shows how to find the monthly FWS evaporation. The section that follows it explains how this procedure was derived.
To find the monthly freewater surface evaporation for Grangeville, the following steps are followed:
1. Find the annual FWS lake evaporation from the map (Grangeville annual FWS evaporation is 30 inches).
2. Determine the region in which Grangeville is located (Grangeville is in Region 2).
3. Multiply the standard curve for Region 2 (Table 1) by the FWS evaporation amount to obtain monthly freewater evaporation in inches for a shallow lake
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Percent
FWS0.5 0.2
1.5 0.4
3 0.9
8 2.4
12 3.6
15 4.5
19 5.7
17 5.1
11 3.3
6 1.8
4 1.2
3 0.9
100 30
You are cautioned that the above values are not pan evaporation rates. If you want a conservative design value, assume zero evaporation for December through February and design the lagoon or storage capacity accordingly. Also be careful about assigning a great deal of credibility to rounding in the decimals.
All maps used in this paper are referenced in Appendix A.INTRODUCTION Reliable estimates of weekly and monthly shallow lake evaporation rates for specific locations in the state of Idaho are required by engineers and others engaged in planning, design and operation of shallow ponds, lagoons, and pesticide wash-water holding areas, as well as by state licensing and regulatory agencies. The design procedures and permit systems require values of the monthly evaporation from the water surface. These values are not easily obtained nor easily computed.
These free water surface (FWS) values are used in a water-balance procedure for the pond (Figure 1). An accurate water-balance between the inflow (precipitation, runoff, wash water, pumping, etc) and outflow(evaporation and overflow) will result in a properly sized pond that is neither too large to be economical nor too small to hold all the required water.
The purpose of this bulletin is to present a simple method for obtaining reliable monthly point location estimates of FWS evaporation in Idaho.DESCRIPTION OF THE METHOD In order to keep the method simple, easy to use and reliable, existing data and maps were used rather than a formula based system. Such a method was developed for Wyoming by Warnaka and Pochop (1988) and Pochop and Athren (1987). Many existing design procedures use formulas to compute FWS evaporation from pan evaporation data. Because of the sparse pan evaporation network, maps are used extensively to determine the pan and FWS evaporation (Kohler and others, 1959; Farnsworth and others, 1982). Since a map of mean annual FWS evaporation is available, the monthly distribution of FWS evaporation as a percentage of the annual FWS evaporation can be determined if the amount of evaporation in any month can be determined. The actual FWS evaporation for any month can then be obtained by applying the derived monthly distribution to the annual total taken from the NOAA evaporation atlas (Farnsworth and others, 1982).
Data
The data consisted of both published and unpublished National Weather Service (NWS) records, reference evapotranspiration (ET) data computed from climatological data and maps published by NOAA. Specifically these were:
1. Long term pan evaporation records for nine stations in Idaho and four stations in surrounding states. These data were subjected to extensive quality control (Appendix D).
2. May through October pan and FWS evaporation and annual FWS evaporation maps for Idaho published in the NOAA evaporation atlas (Farnsworth and others, 1982) (Figures A-1, A-2 and A-3).
3. Mean monthly reference ET (not adjusted for station aridity) for 98 stations in Idaho (Appendix B; Allen and Brockway, 1983).
DATA ANALYSIS The data analysis procedure consisted of two steps. In the first step, the ET and pan evaporation data were used to manually partition the state into regions of similar monthly patterns of evaporation. This was done on the assumption that the patterns of potential ET were similar to those of FWS evaporation. In this step, allowances needed to be made for the heat storage in the shallow ponds that is not present in either the pan or ET evaporation data (Figure 2). During the second step, the ET and pan evaporation data and the NOAA Atlas were used to estimate FWS evaporation curves by month for each region identified in Step One.
Regionalization
A design procedure should be as simple as possible. This required that the state be divided into fewer than ten regions corresponding roughly to NOAA's ten climatic divisions for the state. Each of these should exhibit very similar patterns of monthly evaporation.
Since only nine usable pan evaporation stations were available (only one in northern Idaho), the 98 ET values were used. These consisted of reference crop ET (alfalfa) computed for each month for March through October by the FAO Blaney-Criddle procedure. In using these data, the assumption was made that the percentage of annual ET which occurs in any one month is also an approximation of the percent of the annual FWS evaporation that occurs in that month. This is a relatively good assumption because the computed ET is the potential rate and is sensitive to the same meteorological variables as is the lake or pan evaporation. These values gave the monthly percentages for eight months leaving only four months to be determined by other means.
Monthly ET data for the 98 stations were expressed as a percentage of the May through October total ET since this was the same time period available from the NOAA charts. By inspection, 64 of the 98 ET stations were grouped into six rather than ten geographic regions based on similarities of the percentage of May-October ET which occurred during each of the eight available months.
Discriminant analysis was then used to verify the validity of the manual grouping and to classify the remaining 34 stations. Discriminant analysis develops equations using the known data groupings obtained by inspection. These are then used to place previously ungrouped stations into one of the six groups based on the similarity of the statistical parameters for each station as well as performing a check on the grouping as obtained by inspection. Various combinations of the climatic regions were attempted in order to determine the best dimensionless monthly evaporation curve with the minimum number of regions. This resulted in the five regional groups as shown in Figure A-4.
This map should be used with caution whenever the location in question is near a regional boundary. The user should always consider the fact that the point could also fall into the adjacent region. This is especially true on the north side of the Snake Plain at the boundary between Region 3 and 4.
Annual Percentage Distribution
The monthly percentages were developed by first determining the percentage of the annual FWS evporation that occurs in the May-October period.
The May-October FWS evaporation is obtained from Figure A-2 and the annual FWS evaporation from Figure A-3. From these two values, the ratio of May-October FWS evaporation to the annual value may be obtained. For the entire state, this ratio ranged from 0.71 in the southeast (Region 5) to 0.83 in the north (Region 1). The final values determined for each region were:
Region
Percent of Annual FWS evaporation
May-Oct
Nov-Apr
1
80
20
2
80
20
3
78
22
4
79
21
5
73
27
The actual percentages of the annual total that occurred in each of the months May through October was obtained by using the percentages for these months from both the ET and pan evaporation data.
As an example of this calculation, Parma is in Region 3. The May-October FWS evaporation is 35 inches while the annual value is 45 inches. Multiplying the 35 inches by the monthly percentages developed from the ET values gives the monthly FWS evaporation shown below:
March
April
May
June
July
Aug
Sept
Oct
TOTAL
March-Oct ET
2.72
5.55
7.62
9.35
10.58
8.67
6.45
3.87
54.81
inches
ET in percent of
6
12
16
20
23
19
14
8
118 %
May-Oct
sum of May-October = 46.64 inches or 100 percent
FWS in percent of
4
9
12
16
18
15
11
6
91 %
annual
The May-October period represents 78 percent of the annual FWS total evaporation of 45 inches. Thus if all of the above percentages are multiplied by 0.78 the sum of the May-October percentages is 78 and the sum of the March-October period is 91 percent. Ninety-one percent of the annual FWS evaporation occurs in these eight months leaving nine percent to be distributed over the remaining four months.
By plotting cumulative percentages and assuming that January is the low month at not less than 0.5 percent nor more than one percent, the amounts for these four months can be determined. Figure 3 shows the cumulative percentages for Parma with estimated values for November-Febrary included. The percentages for November through February were estimated simply by extrapolating this cumulative curve. The final annual distribution for Parma is:
Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec FWS in percent of annual
1 1 4 9 12 16 18 15 11 6 5 2 These same steps were repeated for all stations in each Region. At any time any rounding was done, it was towards the following month. This was to account for the the heat storage in the pond by shifting any possible extra evaporation to a later month.
The final step was to tabulate the annual distributions for all stations in each region. By visual inspection and simple averaging, a representative curve of the percentage FWS evaporation for each region was obtained. In regions with no pan evaporation stations, ET values were utilized. The ET monthly percentage values were assumed equal to pan evaporation percentages since this was approximately the case for all locations which had both pan and ET data. The final distributions for all five regions are given in Table 1.
TABLE 1. Monthly shallow lake evaporation percentages for Idaho to be used with the map showing regions of monthly FWS evaporation
Area Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1 1.0 1.5 3 9 12 14 19 18 11 6 4 1.5 2 0.5 1.5 3 8 12 15 19 17 11 6 4 3 3 0.5 1.5 4 8 14 15 17 16 10 6 5 3 4 1.0 2.0 4 7 12 15 19 16 11 6 4 3 5 1.0 3.0 5 10 12 14 16 15 10 6 5 3 DISCUSSION
The ET monthly percentages for each station within each region showed remarkable agreement with each other when expressed as percentage of the sum of May through October values. These percentages were also very similar to those derived from the pan evaporation data within the same region. The results seem to indicate that ET is a consistent parameter for use in dividing the state into regions of similar evaporation activity since the same climatological variables affect both ET and pan evaporation.
The user needs to be cautious in the application of this procedure. Two problems in particular stand out. Whenever regionalization is used, there is always a problem at the boundary between two regions. As an example, Regions 3 and 4 have a boundary running through an area where there are many lagoons. If the lagoon is designed according to one Region or the other, there will be some differences. This is illustrated as follows:
Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec difference
-0.5 -0.5 0 1 2 0 -2 0 -1 0 1 0 This shows that Region 3 balances evaporation in November with evaporation in January and February. More important, the influence of the many stations in the Snake Plains is seen in that Region 3 has more evaporation in the spring months while the higher elevation Region 4 has a higher percentage of its evaporation in July through September. The user should always make the computations for both regions whenever the location in question is near a boundary.
The second major problem has to do with the energy balance of a pond, in particular, the heat storage (Figure 2). In general, as we move from a pond the size of an evaporation pan to one of several acres, the amount of time it takes to heat or cool the water will result in progressively longer lag times for evaporation. Because of this, any ponds larger than one acre in size and more than six feet deep should be designed with possible lags in mind. Thus the percentages for June, July and August could have some amount moved to the following month so that the highest evaporation occurrs in August rather than July. The easiest way to do this is to reverse the percentages for July and August.
The elevation of the design location will also enter into the lagging estimation. The changes in evaporation with elevation were considered when the maps were drawn. However, as elevation increases, there will be a general decrease in air temperature and a slight lag in the time of maximum temperature and evaporation. This is especially true in areas with large amounts of snowcover where the air temperature will not begin to significantly increase with time as fast as in snowfree zones because of the energy necessary to melt the snow.The designer must use any experience gained from working in an area to help interpret these design percentages and amounts. In addition, since the FWS evaporation maps depict mean annual amounts, the results can be adjusted up or down based upon the factor of safety that is required for a particular pond. Because inflows and precipitation cannot be predicted, the designer must exercise prudent judgement in the pond sizing, including the projected evaporation amounts.
CONCLUSIONS Due to the large number of ET stations utilized in this analysis and the relatively close statistical fit between stations in each group, the groups specified should be representative of regions with similar ET activity in Idaho. The nine pan evaporation stations fit these regions very well. Therefore, the regional curves developed here should be representative of monthly shallow pond evaporation rates in the various regions and should be very useful in the planning, design, operation, and regulation of shallow ponds in the state of Idaho.
REFERENCES Allen, R.G. and C. E. Brockway. 1983. Estimating consumptive irrigation requirements for crops in Idaho. Research Technical Completion Report. Idaho Water and Energy Resources Research Institute, University of Idaho.
Farnsworth, Richard K. and Edwin S. Thompson. 1982. Mean Monthly, seasonal, and annual pan evaporation for the United States. USDC, NOAA, National Weather Service. NOAA Technical Report NWS 34.
Farnsworth, Richard K., Edwin S. Thompson and Eugene L. Peck. 1982. Evaporation atlas for the contiguous 48 United States. USDC, NOAA, National Weather Service. NOAA Technical Report NWS 33.
Kohler, M.A., T.J. Nordenson and D.R. Baker. 1959. Evaporation maps for the United States. National Weather Service Technical Paper 37.
Pochop, Larry, John Borrelli, and Victor Hasfurther. 1985. Design characteristics for evaporation ponds in Wyoming. Wyoming Water Resources Research Center. University of Wyoming.
Warnaka, Karen and Larry Pochop. 1988. Analysis of equations for free water evaporation estimates. Water Resources Research 24(7):979-984.
