Development of spatial similarity‐based modelling to improve integrated lake water quality management in Malaysia

Abstract

AbstractThis study describes similarity‐based modelling techniques used to develop a spatially based water quality prediction model to facilitate sustainable lake water quality management. The lentic nature of lakes allows them to slowly absorb pollutants over a long period of time with readily noticeable signs, causing symptoms to appear only when the water quality has significantly degraded, meaning the risk of improper water quality management can be very high. Thus, failure to establish sustainable planning at the watershed scale was found to be a major threat of water quality degradation, from extemporary approaches often practised in lake management. Accordingly, the developed model is tailored for lakes facing moderate to serious water quality data limitation. The geodatabase integrates the identified driving factors of physical, social and water quality with significant influences on the status of lake water quality. A 1 km buffer radius with percentages of built‐up area, population, lake surface area and rainfall is measured. Calculation of a water quality index was then quantified on the basis of a similarity‐modelling technique. Lake Putrajaya was chosen as a control point for developing the indicator index. A total of 93 recreational lakes within Selangor and Kuala Lumpur were selected as modelling points. The results of this study indicated the similarity technique of spatial modelling is sufficiently reliable to be applied as an early assessment indicator. From the 93 lakes in this study, none feel in the category of either “bad” or “excellent,” with the majority being in class 3 (medium water quality status) and only four considered as having a good water quality condition. The balance of 35 lakes was considered to exhibit poor water quality. The model output is an indicator index, providing classification guidelines for the water quality status of the assessed lake as an early assessment tools.