Water resources engineering is an important factor when discussing social, economic and environmental concerns. Our student Brandon Hanson makes this connection in Bangkok, Thailand. This case study was reported in by the Asian Institute of Technology in 2011, under the name Identifying Prominent Explanatory Variables for Water Demand Prediction Using Artificial Neural Networks: A Case Study of Bangkok, by Mukand Singh Babel and Victor R. Shinde, and brings up concerns of predicting water demand in the city of Bangkok. This is a believable source as it describes how the Artificial Neural Network System models were developed to predict short term and long term water demands for Thailand as another article describing possible methods of evaluating artificial neural network techniques for municipal water demand here.
This study relates to water resources engineering tackling one of the big questions when managing water, predicting how much water do we need for an urban environment. A method for predicting water demand accurately can allow engineers to better ration water for periods of emergency like fire or drought. This is very important for a region like Thailand where the rainy season bring heavy rain that causes flooding and the dry season brings extensively long periods of drought. It provides a system of predicting the water demand for a city based on climate variables such as rainfall and temperature to predict short term water demands. However it was observed in the study that as the prediction time for the water demand increase, the accuracy with using only climate variables decreases. Longer periods of prediction required a combination of variables, including population, number of household connections, education status, per capita GPP, maximum temperature, rainfall and relative humidity. Due to Bangkok being located in a particular tropical region, the relative humidity played a particularly important factor. They could go into more details on the limitation of their system, they mention that their model is is particularly useful in generalization problems but not how this generalization can affect it’s accuracy for short term predictions.(Babel, 2010)
Thailand’s changing climate has been contributing to the instability and challenges. The Central plains of Thailand have no water reservoir of their own and must rely on the dams in the northern region of the country to supply them with water. Due to the long periods of drought each year has led to the decrease in the water flowing through the dams in Thailand. For the past few years Bangkok, as well as the surrounding country, has been affected by the most severe water shortage the country has had in the past two decades, with water rationing being imposed in some provinces and hotels told to minimize their laundry loads.(Suwal, 2007) This system for predicting water demand is both a social and economic concern, allowing people to predict the cost of delivering water as well as determining if there is enough water available for the residents. The short term models handles environmental concerns by scheduling pump operations and reducing detention time in storage tanks, improving the water quality.
Suwal, S. (2007). Water In Crisis – Spotlight Thailand. Retrieved February 14, 2017, from https://thewaterproject.org/water-crisis/water-in-crisis-thailand
Babel, M., & Shinde, V. (2010, December 3). Identifying Prominent Explanatory Variables for Water Demand Prediction Using Artificial Neural Networks: A Case Study of Bangkok. Water Resources Management, 25(6), 1653-1676.
Fernquest, J. (2015, February 6). Water shortages nationwide as annual drought begins. Retrieved February 14, 2017, from http://www.bangkokpost.com/learning/learning-news/467401/water-shortages-bite-as-annual-drought-sets-in