Prediction of dissolved oxygen, biochemical oxygen demand, and chemical oxygen demand using hydrometeorological variables: case study of Selangor River, Malaysia

Salih S. Q., Alakili I., Beyaztas U., Shahid S., Yaseen Z. M.

ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY, cilt.23, sa.5, ss.8027-8046, 2021 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 23 Sayı: 5
  • Basım Tarihi: 2021
  • Doi Numarası: 10.1007/s10668-020-00927-3
  • Dergi Adı: ENVIRONMENT DEVELOPMENT AND SUSTAINABILITY
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, International Bibliography of Social Sciences, PASCAL, ABI/INFORM, Agricultural & Environmental Science Database, Aquatic Science & Fisheries Abstracts (ASFA), BIOSIS, Business Source Elite, Business Source Premier, CAB Abstracts, Geobase, Greenfile, Index Islamicus, Pollution Abstracts, Veterinary Science Database, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.8027-8046
  • Anahtar Kelimeler: Water quality parameters, Prediction, Hydrometeorological variables, Tropical environment, WATER-QUALITY INDEX, VECTOR AUTOREGRESSION, MODEL, FORECAST, EVALUATE, PAHANG
  • Marmara Üniversitesi Adresli: Evet

Özet

In this research, three water quality (WQ) indexes, namely dissolved oxygen (DO), biochemical oxygen demand (BOD), and chemical oxygen demand (COD), in Selangor River of peninsular Malaysia were simulated using a stochastic model based on vector autoregression (VAR). The simulation was adopted based on three modeling scenarios of inputs as predictor: (i) related WQ parameters, (ii) WQ parameters and river flow data, and (iii) WQ parameters and rainfall data. The WQ parameters as input were determined based on the correlation analysis. The numerical analyses revealed that the prediction accuracy of VAR model substantially increases with the increase in input number. The model provided better accuracy in predictions of WQ indexes (root mean square error approximate to 0.11 and mean absolute error approximate to 0.26) when all environmental, hydrological, and climatological variables were considered. Further improvement in model performance (root mean square error approximate to 0.0248 and mean absolute error approximate to 0.1259) can be achieved if physiochemical parameters like suspended solid material and the turbidity are used as additional inputs.