## Info

Stresses due to dry or wet conditions and/or high salinity concentrations may reduce The potential root water extraction rate. The water stress in SWAP is described by the function proposed by Feddes et al. (1978), which comprises a coefficient that takes values between zero and one. Under conditions wetter than a certain "anaerobiosis point" (hi) water uptake by roots is zero, as well as the coefficient a. Likewise, under conditions drier than "wilting point" (h4), a is also zero. Water uptake by the roots is assumed to be maximal when the soil water pressure-head is between h2 and h3 and hence a-value is one in that case. The values of a decrease linearly with h for h values lower than h4 but larger than h3. According to Leenhardt et al. (1995), the Feddes et al. (1978) root water-uptake model has the advantage that not only considers the crop transpiration reduction due to lower soil-water contents, but also takes in account the negative effect of water excess in the soil root zone. Utset et al. (2000) showed that the Feddes et al. (1978) model fits better to actual data, obtained in tropical conditions, than the model provided by Van Genuchten et al. (1987). This SWAP feature could be very important, since flooding could be regionally more severe in the future (IPCC, 2000, Rosenzweig et al., 2002; Utset et al., 2006).

For salinity stress the response function of Maas and Hoffman (1977) is used in SWAP (Van Dam, 2000), as this function has been calibrated for many crops.

Besides calculating crop yields through the WOFOST module, a simpler approach to calculate yield reduction as function of growing stage can be used in SWAP (Doorenbos and Kassam, 1979; Smith, 1992). The ratio between actual to potential transpirations is known as "Relative transpiration ratio" (Van Dam, 2000). The relative transpiration reductions can be related to the effects of water stress (De Wit et al., 1978; Van Dam 2000). The relative yield of the entire growing season is calculated as product of the relative yields of each growing stage.

Two different types of irrigation can be specified in SWAP (Kroes et al., 2002). Either a fixed irrigation can be specified, or an irrigation schedule can be calculated for a specific crop according to a number of criteria. A combination of fixed and calculated irrigations is also possible. An example of this is a fixed irrigation (preparation of the seed bed) before planting and calculated irrigations based on soil moisture conditions after planting. Fixed irrigations can be applied the whole year. Irrigation scheduling can only be active during a cropping period. The irrigation type can be specified as a sprinkling or surface irrigation. In case of sprinkling irrigation, interception will be calculated (Kroes et al., 2002).

3.2.3 SWAP Model. The Simple Approach to Estimate Crop Water-Use

Kroes and Van Dam (2003) described the SWAP performance. According to them, the simple SWAP crop-growth approach represents a green canopy that intercepts precipitation, transpires and shades the ground. Leaf area index, crop height and rooting depth must be specified as functions of the development stage. SWAP simulates crop growing on a daily basis. The development stage (DVS) at a given day j depends on the development stage at the previous day and the daily temperature, according to:

sum where Tsum is the required temperature sum and Teff is the effective daily temperature, calculated from mean daily air temperature minus a minimum starting temperature (3 °C). DVS reaches 1 at anthesis and 2 at maturity, according to the temperature sums at these two stages.

Van Dam (2000) provided the theoretical SWAP background. SWAP solves the Richards equation numerically, subject to specified initial and boundary conditions and the hydraulic functions of the soil. The maximum root water extraction rate (Sp) at a depth z, considering a uniform root-length density distribution, can be calculated from:

Droot

Where Droot is the root density fraction, integrated over the rooting length density at this depth and Tp is the potential transpiration rate, which is subject to atmospheric conditions.

Actual root water uptake can be estimated through:

The a values range from zero (no root-water uptake) to one (maximum water-uptake, no stress) according to the Feddes et al. (1978) function. These values change according to actual soil water-content, but the function is crop-dependent (Van Dam, 2000). Utset et al. (2000) showed that the original parameters of the Feddes et al. (1978) model can be used to simulate potato water-use in conditions that are very different from those existing where the model was originally applied. Therefore, the parameters applied to sugarbeet in the Netherlands have been used in this assessment.

The potential reference evapotranspiration (ETP) is calculated in SWAP by the Penman-Monteith approach, although the user may introduce other ETP calculations (Van Dam, 2000). In addition, the crop coefficients Kc must be introduced to convert reference ETP on maximum crop evapotranspiration.

SWAP first of all separates the potential plant transpiration rate Tp and potential soil evaporation rate Ep and then calculates the reduction of potential plant transpiration and soil evaporation rates. The soil evaporative component can be separated from the total evapotranspiration calculated by the equation:

where m is a crop-dependent coefficient and I is the Leaf Area Index. The transpirative component is therefore calculated from the difference between the total evapotranspiration and the evaporative component. Ritchie (1972) and Feddes (1978) used m = 0.39 for common crops.

Secondly, actual transpiration is calculated considering the root water uptake reduction due to water stress, using equation [2]. Actual soil evaporation can be calculated from the Maximum Darcy flow at the soil surface, which depends on actual soil water content and hydraulic conductivity, although several other options for calculating actual soil evaporation are implemented in SWAP (Van Dam, 2000).

The simple SWAP crop-growth approach does not estimate the crop yield. However, the user can define yield response factors (Doorenbos and Kassam, 1979; Smith, 1992) for various growing stages as functions of the development stage. During each growing stage k, the actual yield Yak (kg ha-1) relative to the potential yield Ypk (kg ha-1) during the said growing stage is calculated by:

## Post a comment