Research On Water Market Trading Mechanism Based On Spatiotemporal Characteristics

Water markets have been rapidly developed over the last few decades.Compared to the water markets in U.S.and Australia,China's water markets are small and still have many problems to solve.This study analyses several problems in water markets and water trading.First,understanding the situation of a water market relies on a reasonable assessment framework.This study proposes a water market assessment framework based on the history and experience of Xiying Irrigation in the Shiyang River Basin,northwest of China.This framework includes overall and network assessments,and centrality in network theory is introduced to reflect the spatial structure.At the same time,this study compares the Australian water market,and points out that the further development of the water market in Xiying Irrigation still faces some obstacles.The water market in Xiying Irrigation needs to increase the incentives for water trading and reduce administrative control.Furthermore,to increase the efficiency of water markets,this study proposes a two-phase model,which can provide optimal assignments to maximize social welfare and recommend appropriate trading price profiles.The solution also has a good incentive property,i.e.,participants are willing to follow the trade assignments given the pricing plan.Based on the two-phase model,this study introduces three objective functions for price setting,aiming at price stability.According to the simulation based on historical trading data in Xiying Irrigation,it is concluded that the “threshold balance” objective function is preferred by more traders in the market than the other two objective functions.This two-phase model can increase social welfare more than pooling method or pooling method for canals,given the condition that there are trading restrictions.Furthermore,administrators and participants typically want to get prediction information on future prices to adjust policies or irrigation plan accordingly.However,price prediction is not that straightforward because it depends on weather,location,policy and many other factors.This study uses Australia's water trading data from 2008 to 2017 to build price prediction models.Boosted decision tree and three additional models are applied to address this problem.Among them,boosted decision tree provides the best result.These results hopefully can help participants and policy makers make better decisions and improve the activity of water markets.To sum up,this study systematically analyzed the assessment method,trade matching,price setting,and price prediction in water markets.Hopefully,these results can help understand and build water markets in China.