By Sun Yingbao (China)
Climate change is having serious impacts on the world’s water systems through more flooding and droughts. Although Costa Rica is a water-abundant country, the extreme droughts and extreme rainfall pattern caused by climate change have undermined the diverse vegetations in Costa Rica. One of the common approaches of mitigating the influence from extreme droughts and rainfall pattern on vegetations is to build up and also intensify the irrigation system. An irrigation system is composed of the inlets, the outlets as well as pipes. This thesis proposes an automated model to help design and optimize the irrigation system under several geo-spatial constraints. A weighted cost function was defined to calculate the cost of construction of the irrigation system. And this cost function was also the target to be minimized during optimization process. Two cost costs required the input from user. The model developed in this thesis optimized the irrigation system based on the Steiner minimum tree algorithm. The resulting layout of the optimized irrigation system is a Steiner-tree like network which connects all inlets and outlets with minimum cost under several geo-spatial constraints. To assess the sensitivity of the model to the changes in model parameter settings, the optimization process was tested in 4 scenarios with different model parameter settings. The sensitivity of model to changes in parameter settings was measured by symmetric Hausdorff distances which is used to measure the similarity between two spatial objects.
Keywords：irrigation system; optimization; cost function; Steiner minimum tree; sensitivity analysis; symmetric Hausdorff distance.