Impacts from Climate Change

Modelling framework

Three different types of simulation models are identified to be necessary for a proper assessment and exploration of adaptation strategies: rainfall-runoff simulation, water allocation programming and crop production prediction. Artificial neural networks (ANNs), ZWAM and SWAP (van Dam et al., 1997) have been selected as the appropriate models to be used.

Rainfall-runoff modelling is required to analyse the processes leading from rainfall to runoff that can eventually be used for irrigation purposes. Several approaches exist, ranging from more physically based methods using semi-distributed hydrologi-cal models to simplified rainfall-runoff statistical regression models. We have selected here to use an artificial neural network approach, since this was already developed, well-tested and validated for similar simulations (Dawson and Wilby, 1998; Sanjikumar and Thandaveswara, 1999; ASCE, 2000).

An ANN can be described as an information processing system that is composed of many non-linear and densely interconnected processing elements or neurones. ANNs have the ability to extract patterns in phenomena and overcome difficulties due to the selection of a model form such as linear, power or polynomial.

An ANN algorithm is capable of modelling the rainfall—runoff relation due to its ability to generalize patterns in noisy and ambiguous input data and to synthesize a complex model without prior knowledge (Dawson and Wilby, 1998; Coulibaly et al., 2000).

For rainfall-runoff simulation, 31 years (1972-2002) of records from the Chadegan Dam gauging station that measures total inputs to the dam (upper catchments and transferred water from Tunnel No. 1 and Tunnel No. 2) have been used to train and test the model. Applying different inputs, ANNs models and architectures, the recursive Elman Networks with 7-2-1 architecture was found to be suitable for the study area.

Water allocation between and within different sectors is of paramount importance in Zayandeh Rud. The basin is highly developed in terms of water resources and any change in water allocation has a direct impact on other water users. To deal with these issues, the Zayandeh Rud Water Allocation Model (ZWAM) was developed for this study. The model is able to simulate different water allocation policies, dam operations, environmental issues and examine different scenarios for future changes in the study area. The model is node oriented and the main water demand sites along the river have been embedded in the model. ZWAM is based on similar concepts to the WEAP model (WEAP, 2002).

The agro-hydrological analyses at field scale have been done using the Soil-Water-Atmosphere-Plant (SWAP) model. The model is a physically based one for simulating water, heat and solute transfer in the saturated and unsaturated zone. The model is also capable of simulating crop growth using meteorological data, irrigation planning and phonological crop data. A more detailed description of the model can be found in van Dam et al. (1997).

Indicators

To describe the current state of water resources in the basin as well as its future status, a number of indicators have been selected that describe the state of the environment (mainly wetlands) and food security (Aerts et al., 2003).

The indicators that can quantify the state of food security are water allocated to agriculture (m3/year) food production (tonnes/year) and crop-derived energy production (calories/year). The last indicator makes it possible to compare food production from different crops. Furthermore, since environmental quality is mainly a function of water availability and amount of water that reaches the swamp, it has been decided to use three environmental indicators as 'years with inflows < 75 X 106 m3', 'years with inflows > 75 X 106 m3 and < 140 X 106 m3' and 'years with inflows > 140 X 106 m3'.

In summary, the following indicators are used to express the current state and the expected state in the future with and without adaptation strategies:

• water allocated for food production (m3/year);

• total food production (tonnes/ year); and

• total food energy production (kcal/year).

Fig. 6.3. Mean monthly inflows in the main reservoir according to the A2 (top) and B2 (bottom) scenarios.

Impacts of low-flow years on food and environment

The impact of climate change on water, food, industry and environment has been assessed by using the modelling framework as described above. Note that impacts alone without explicit adaptation can be considered the 'business as usual' adaptation strategy.

Besides the impact of climate change also other expected changes, such as population growth, increasing domestic water demands, increasing food requirements and growth in industrial water demands, have been included. A few other drivers are not included, of which the most important one is technological innovation, including the introduction of crop species that are more resistant to water shortage, water salinity or crops that are high-yielding.

Overall, the analysis shows that the basin is under threat of climate change (Table 6.2). The change in average basin rainfall and the temperature increases lead to a

9 10 11

No. of consecutive dry years

Fig. 6.4. Frequency of number of successive dry years for historical and climate change periods.

reduction in water resources quantity and quality. With respect to the future population growth, the basin's domestic water requirement will reach 344 and 540 X 106 m3 in 2039 and 2099, respectively (about 150 X 106 m3 at present). The growing rate of industrial water demand has been assumed to be 1% up to 2010, resulting in about 115 X 106 m3 by 2010. It is assumed to remain constant after 2010. This assumption is based on the water conservative policies for industry.

Using the trained ANNs model, the streamflows have been simulated for the selected 2010-2039 and 2070-2099 periods for both the A2 and B2 scenarios. The mean monthly flows and their distribution under the A2 scenario show significant changes in timing and volume compared to the historical flows (Fig. 6.3). But, the changes under the B2 scenario are lower and stream flows show almost the same temporal pattern. The sequences of successive dry and wet years have been estimated and are shown in Fig. 6.4. From this figure it is clear that the maximum number of successive dry years during the observed period is 2 years, and increases to 11 years for the A2 scenario and to 3 years for the B2 scenario.

Considering the changes in rainfall and river flows under the climate change scenarios, estimates were made for the groundwater budget for the periods 2010-2039 and 2070-2099. Results show that for most sub-basins an overdraft can be expected. For instance, Najaf Abad aquifer may experience a 2.6 m/year lowering of the water table. The analysis indicates clearly that an increase in exploitation of groundwater is not possible at all, unless new sources (e.g. transfer of water from adjacent basins) for recharge are to be implemented. In general, scenario B2 for the period 2010-2039 shows a declining trend in groundwater level, and this was even more profound for the A2 scenario for the period 2070-2099.

Climate change impacts on agricultural production are the result of two processes, increasing air temperature (i.e. increasing transpiration) and CO2 enrichment of the atmosphere. These two have been separately considered in this study. In the first step, the SWAP model has been run for the staple crops including wheat, barley, rice and potatoes, and the current climate situation as well as the selected future periods, with scenarios A2 and B2. The general pattern is that considering solely the

variation in temperature and rainfall, there is not much change in relative yield for the period 2010-2039 compared with the current situation. However, for the period 2070-2099, 10-15% reduction in crop yields can be expected.

Contribution of CO2 on crop yields was considered next. It is expected that CO2 levels in 2100 will reach 640 ppm, which means an increase of more than 300 ppm compared to the present situation (Parry et al., 1999; IPCC, 2000). Responses of plants to this phenomenon will be different. It is expected that C3 plants, such as wheat, barley, rice and potatoes, will respond more positively to rising levels of CO2 (see also Chapter 3). In a case study in Tabriz (north-west Iran), Koochaki and Mahallati (2001) also reported an increase in crop yields due to CO2 enrichment. Based on the above studies, it has been assumed that the combined changes of rainfall and CO2 enrichment will increase crop yields up to 25% in the study area. However, this is based on field-scale analysis and it should be emphasized that this is only potential yield. In addition to potential yield, the actual yield is a function of irrigation water quality and quantity that are anticipated to decline on a basin-level scale in the future.

0 0

Post a comment