Modelling framework

Climate change and adaptation strategies should be evaluated on different spatial scales. Since food production is a critical issue in Walawe, the field-scale model SWAP (van Dam et al., 1997) has been used to deal with water and food issues. To consider these field-oriented analyses in a broader context and to include environmental issues, the basin-scale model WSBM (Droogers et al., 2001a) has been set up for the basin.

The SWAP (Soil-Water-Atmosphere-Plant) is a one-dimensional physically based model for water, heat and solute transport in the saturated and unsaturated zones, and also includes modules for simulating irrigation practices and crop growth. The water transport module in SWAP is based on the well-known Richards' equation, which is a combination of Darcy's law and the continuity equation. A finite difference solution scheme is used to solve Richards' equation. A detailed description of the model and all its components is beyond the scope of this chapter, but can be found in van Dam et al. (1997).

In terms of irrigation requirements and food production, rice is the most important crop in Walawe Basin. It not only provides food for the people in the basin, but also is an important crop for export to the capital Colombo. Normally two crops can be grown during the year, one at the Maha and one at the Yala season.

Since no detailed crop characteristics were available, crop yields were computed using a simple crop-growth algorithm based on Doorenbos and Kassam (1979). The basic assumption of the simplified crop production function is that actual yield is a function of potential yields and water stress:

where Y .. and Y .. are the potential and actual yield for a specific year i, and T .. and pot,i act,i l / r J y pot,i

Tact i the potential and actual transpiration for year i. Sometimes evapotranspiration is considered instead of just transpiration, since determination of crop transpiration alone is difficult. Doorenbos and Kassam (1979) expanded this approach by including the effect that the sensitivity of the crop to water stress during subsequent growing periods is not constant:

where K is yield reduction factor (—) indicating whether a crop is sensitive (>1) or less sensitive (<1) to water stress. K can have different values for different growing periods y.

A main drawback of this approach is the tendency to estimate high yields during low Tpot (hence low solar radiation) periods, and vice versa. We have therefore followed the approach to adjust the Ypot accordingly to this 7^, as applied successfully previous (Droogers et al., 2001b). The potential yield for a certain year is assumed to be a linear function of the real maximum potential yield as obtained during very favourable climate conditions and optimal farm management:

pot,max pot,i pot,max where Y and T are the maximum crop yield and maximum transpiration pot,max pot,max l / r during the period of 30 years as considered in this study.

Obviously, the option to use a detailed crop modelling approach would be preferred, but since detailed crop parameters were lacking we have used the simplified approach as indicated. It should be emphasized here that this simplified approach is used very often and has proven to provide reasonable estimates (Droogers and van Dam, 2004).

Crop production is affected by the air's CO2 level. Photosynthetically active radiation (PAR) is used by the plant as energy in the photosynthesis process to convert CO2 into biomass. It is important in this process to make a distinction between C3 and C4 plants (see also Chapter 3). A large number of experiments have been carried out over recent decades studying the impact of increased CO2 levels on crop growth. The Center for the Study of Carbon Dioxide and Global Change in Tempe, Arizona ( has collected and combined results from such experiments for different crops, including rice (Oryza sativa L.). According to 26 references for rice, average biomass increase was 31% and 47% for increases in CO2 air concentrations of 300 and 600 ppm, respectively.

At the basin scale, the Water and Salinity Basin Model (WSBM) was set up to link field-scale issues to the entire basin and to connect food issues with environmental ones. The main objective of WSBM was to create a simple and transparent water accounting model, to be used for quick analysis of river basin processes (Droogers et al., 2001a). The model focuses on extractions for irrigation and the associated return flows from these systems. The model also includes a simplified urban and industrial water extraction component. WSBM works in an object-oriented style and was set up in Microsoft Excel to support transparency and flexibility. The concepts of WSBM are similar to that of WEAP (WEAP, 2002), but to ease coupling with the field-scale model SWAP, we have elected to use WSBM instead of WEAP.

WSBM assumes that the river is divided into nodes with a reach defined between two successive nodes. Nodes are located at typical points in the river where stream gauges are present or output is required. Water extractions, or supplies, occur only in the reaches. Using this approach water flows along the river can be simulated by subtracting extractions, or adding supplies, from one node to get the value for the next node. As mentioned before, extractions are defined for urban, industrial and irrigation supplies. For both types of extraction the amount of water and the return flow as a percentage of the extraction must be specified. Obviously, values can be either real data or hypothetical values to explore the effects of different interactions. The whole model was set up to run with a monthly time-step and it was assumed that the response time of the basin was within 1 month, so no time lag in water between months occurs. The model itself was applied, tested, calibrated and validated successfully for a basin in Iran (Droogers et al., 2001a).

The two models were linked where changes in basin hydrology and field-scale hydrology interact. The entire set up is a flexible system where changes and adaptation strategies can be evaluated directly.

0 0

Post a comment