This chapter describes procedures for predicting ETc during non-growing periods. Non-growing periods are defined as periods during which no agricultural crop has been planted. In temperate climates, non-growing periods may include periods of frost and continuously frozen conditions.
Bare soil
Surface covered with dead vegetation
Surface covered with live vegetation
Frozen or snow covered surfaces
The type and condition of the ground surface during non-growing periods dictates the range for ETc. Where the surface is bare soil, then the Kc will be quite similar to the Kc ini predicted in Chapter 6. Where the surface is covered by nearly dead vegetation or some type of organic mulch or crop residue, then the Kc will be similar to that for agriculture that uses a surface mulch. Where the surface is covered by weed growth or growth of 'volunteer' plants, then the Kc will vary according to the leaf area or fraction of ground covered by the vegetation and by the availability of soil water. Where the surface is snow covered or frozen, then the Kc is difficult to predict and a constant value for ETc may have to be assumed.
Single crop coefficient
Where the ground is left mostly bare following harvest, then the Kc following harvest will be strongly influenced by the frequency and amount of precipitation. Kc for bare soil can be calculated as Kc = Kc ini where Kc ini is calculated using the procedure of Chapter 6.
Dual crop coefficient
Where a daily soil water balance can be applied, the user may elect to apply the dual Kc approach of Chapter 7. In this situation, the topsoil layer may dry to very low water contents during periods having no precipitation. Therefore, the values for Kcb and for Kc min in Equations 71 and 76 should be set equal to zero. This provides for the opportunity to predict ETc = 0 during long periods having no rainfall. This is necessary to preserve the water balance of the evaporation layer and of the root zone in total. The daily water balance calculation, given Kcb = 0, will provide the most accurate estimates of ETc during the non-growing periods.
Single crop coefficient
Where the ground surface has a plant residue or other dead organic mulch cover, or where part of the unharvested crop remains suspended above the surface in a dead or senesced condition, then the surface will respond similarly to a surface covered by mulch. In this case, Kc can be set equal to Kc ini as predicted from figures 29 and 30, but the value for Kc ini can be reduced by about 5% for each 10% of soil surface that is effectively covered by an organic mulch.
Dual crop coefficient
Evaporation from dead, wet vegetation can be substantial for a few days following a precipitation event. Therefore, in the dual Kc approach, the value for fc should be set equal to zero to reflect the lack of green cover and fw should be set equal to 1.0 to reflect the wetting of both soil and mulch cover by precipitation.
The dead mulch or vegetation will dry more quickly than would the underlying soil if it were exposed. In addition, the soil will be protected somewhat from evaporation by the dead mulch or vegetation cover. Therefore, total evaporation losses will be less than the TEW predicted from Equation 73. This can be accounted for by reducing the value for TEW by 5% for each 10% of soil surface that is effectively covered by an organic mulch. The value for REW should be limited to less than or equal to that for TEW.
During frost-free periods following harvest, weeds may begin to germinate and grow. This vegetation is supplied with water from storage in the soil profile and from any rainfall. In addition, crop seed lost during harvest may germinate following rainfall events and add to the ground cover. The amount of ground surface covered by vegetation will depend on the severity of weed infestation; the density of the volunteer crop; tile frequency and extent of soil tillage; the availability of soil water or rain, and any damage by frost.
The value for Kcb during the non-growing period can be predicted over time according to the amount of vegetation covering the surface. This can be done through estimates of LAI using Equation 97 or estimates of the fraction of ground cover, fc, using Equation 98.
Single crop coefficient
In the single crop coefficient approach, the value for Kcb determined using procedures in Chapter 9 can be converted into an equivalent Kc by adding 0.05 to 0.15 according to the frequency of rainfall and surface wetting.
It is important that the Kc for vegetation during the non-growing period be limited according to die amount of soil water available to supply evapotranspiration. Otherwise, the law of conservation of mass will be violated. Under all conditions, the integration of Kc ETo over the course of the non-growing period cannot exceed the sum of the precipitation occurring during the period plus any residual soil water in the root zone following harvest that can be depleted by the subsequent vegetation. The root zone in this case is the root zone for the weed or volunteer crops. A daily soil water balance may provide for the best estimate of soil water induced stress and associated reduction in Kc and ETc.
Dual crop coefficient
Under the dual crop coefficient approach, Kcb can be predicted according to the amount of surface that is covered by vegetation using Equation 97 or 98. Then, a full daily soil water balance of the topsoil together with a full daily soil water balance of the root zone can be employed as described in Chapter 7. The soil water balances will automatically adhere to the law of conservation of mass, so that total ETc from the weed or volunteer vegetation will not be overestimated. Again, because the topsoil layer may dry to below wilting point under conditions of sparse rainfall, the values for Kcb and Kc min used in Equations 71 and 76 should be set equal to zero. In this manner, the daily soil water balance with dual Kc calculations can progress throughout the non-growing period with good results.
Where the ground surface is snow covered or frozen, any vegetation will be largely non-responsive and non-contributing to ETc, and the amount of ETc will be closely related to the availability of free water at the surface and to the albedo of the surface.
The albedo of snow covered surfaces can range from 0.40 for old, dirty snow cover to 0.90 for fresh, dry snow. Therefore, the ETc for snow cover will be less than ETo for grass, as 25-85% less shortwave energy is available. In addition, some energy must be used to melt the snow before evaporation.
The use of ETo under such conditions is of limited value, as the assumption of conditions sustaining a green grass cover is violated. It is even possible to obtain negative values for ETo on some winter days where the longwave radiation from the surface is large and the vapour pressure deficit is small. It is under these conditions that net condensation of water from the atmosphere is possible. This would be similar to negative evaporation.
Given the limited value of ETo (or even ETp) under snow covered or frozen conditions, a single, average value may be best used to predict ETc. Wright (1993) found that ETc averaged 1 mm/day over winter periods at Kimberly, Idaho, the United States, that were six months long (1 October to 30 March). The latitude of Kimberly is 42°N and the elevation is about 1200 m. Over the six-year study period, the ground was 50% covered by snow for 25% of the time from 1 October to 30 March. The ground, when exposed, was frozen about 50% of the time. The Kc averaged 0.25 during periods when the soil was not frozen but where frosts were occurring (October and early November). When the ground had 50% snow cover or greater, the ETc averaged only 0.4 mm/day. Wright found that over the six-month non-growing period, total cumulative ETc exceeded precipitation by about 50 mm.
Figure 47 shows the mean measurements of ETc during the 1985-1991 study period. The measurements have high correspondence to the total shortwave radiation energy available on a clear day, Rso, estimated as 0.75 Ra. There is some lag between ETc and Rso and Rs caused by cooler temperatures in January - March as compared to the October - December period. The ETc/Rso ratio averaged only 0.17 over the six-month period, and averaged 0.11 from 1 Dec. - 10 Mar. The ETc/Rs ratio averaged 0.23 over the six-month period, and averaged 0.15 from 1 Dec. - 10 Mar.
FIGURE 47. Mean evapotranspiration measured during non-growing, winter periods at Kimberly, Idaho, United States by Wright (1993)
A similar study conducted in Logan, Utah, the United States (latitude 41.6°N, elevation 1350 m) over an eight-year period showed that ETc varied widely with soil surface wetness and air temperature during the winter months. The 'average' Kc from November to March was 0.5 for days having no snow cover. For days with snow cover, 'Etc' ranged from 0 to 1.5 mm/day. Similarly, Kc is about 0.4 for winter wheat during frozen periods in the region of northern China (latitude near 39°N.
Single Crop Coefficient
The above procedure can provide estimates for the single Kc during non-growing season periods having snow cover or freezing conditions. However, the actual value for Kc is known to vary widely and will be less when water is less available at the soil surface.
Dual Crop Coefficient
A daily soil water balance using the dual crop coefficient approach is necessary to accurately predict ETc under freezing and snow cover conditions. In the dual crop coefficient method, a daily water balance is conducted for the topsoil and the estimate for Kc can be reduced according to available water. However, in addition to the limited validity of the concept of ETo under frozen or snow covered conditions, the evaporation coefficient, Ke, may be reduced when the ground surface is frozen, as the water in a frozen state is less available.
Other, more complex models for predicting ETc under non-growing season conditions, snow cover, and freezing, are available in the literature and should be consulted and perhaps applied when precise estimates for ETc are needed. Some of these are listed in section K of the References.
