Dr.Rasha Fadhil, K. Unami, Md Kamal Rowshon, A. Fikri, W. Aimrun
{"title":"Identification of nominal release policies implemented in a multi-purpose water reservoir","authors":"Dr.Rasha Fadhil, K. Unami, Md Kamal Rowshon, A. Fikri, W. Aimrun","doi":"10.26491/mhwm/154592","DOIUrl":null,"url":null,"abstract":"A water reservoir’s operation should follow a rational policy to ensure adequate water provision for different purposes without adverse effects. However, it is not well-studied how to identify operational policies currently being implemented. This study establishes a new approach to identifying nominal release policies as implemented in a multi-purpose water reservoir. We chose Bukit Merah Reservoir (BMR), located in Perak State, Malaysia, as a study site to examine its release policies for meeting irrigation, municipal, and industrial water demands and for mitigating floods and environmental hazards. The operator of BMR releases the reservoir’s water into two primary irrigation canals, the Main Canal and the Selinsing Canal. Generalized additive models (GAMs) are applied to time series data observed at BMR to identify the annual dynamics of its water management. Operational policies for the release discharges into the two primary irrigation canals are assumed to be based on information on the time-of-year and the reservoir water level. First, a backfitting algorithm identifies each contributing function of the GAMs representing the release policies. Then, spurious oscillations in the functions are removed by total variation (TV) regularization (TVR) to obtain nominal release policies, which are quite reasonable in the sense of conventional reservoir management practice. Finally, the identified nominal release policies are utilized to examine shifts in the operation of BMR during the period from 2000 through 2011. The decomposition of release policies illustrates the two aspects of the irrigation demand’s annual patterns and the hydraulic structures’ functions. The spurious oscillations removed by TVR are considered to represent indecision by the reservoir operator.","PeriodicalId":42852,"journal":{"name":"Meteorology Hydrology and Water Management-Research and Operational Applications","volume":"1 1","pages":""},"PeriodicalIF":0.8000,"publicationDate":"2022-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Meteorology Hydrology and Water Management-Research and Operational Applications","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.26491/mhwm/154592","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"WATER RESOURCES","Score":null,"Total":0}
引用次数: 0
Abstract
A water reservoir’s operation should follow a rational policy to ensure adequate water provision for different purposes without adverse effects. However, it is not well-studied how to identify operational policies currently being implemented. This study establishes a new approach to identifying nominal release policies as implemented in a multi-purpose water reservoir. We chose Bukit Merah Reservoir (BMR), located in Perak State, Malaysia, as a study site to examine its release policies for meeting irrigation, municipal, and industrial water demands and for mitigating floods and environmental hazards. The operator of BMR releases the reservoir’s water into two primary irrigation canals, the Main Canal and the Selinsing Canal. Generalized additive models (GAMs) are applied to time series data observed at BMR to identify the annual dynamics of its water management. Operational policies for the release discharges into the two primary irrigation canals are assumed to be based on information on the time-of-year and the reservoir water level. First, a backfitting algorithm identifies each contributing function of the GAMs representing the release policies. Then, spurious oscillations in the functions are removed by total variation (TV) regularization (TVR) to obtain nominal release policies, which are quite reasonable in the sense of conventional reservoir management practice. Finally, the identified nominal release policies are utilized to examine shifts in the operation of BMR during the period from 2000 through 2011. The decomposition of release policies illustrates the two aspects of the irrigation demand’s annual patterns and the hydraulic structures’ functions. The spurious oscillations removed by TVR are considered to represent indecision by the reservoir operator.