Investigating MODIS-Derived Vegetation Metrics on the HowLeaky Model to Predict Plant Available Water in a Data-Scarce Humid Subtropic Paddock of Australia
Article
| Article Title | Investigating MODIS-Derived Vegetation Metrics on the HowLeaky Model to Predict Plant Available Water in a Data-Scarce Humid Subtropic Paddock of Australia |
|---|---|
| ERA Journal ID | 18556 |
| Article Category | Article |
| Authors | Rawat (Student), M., Nguyen-Huy, T., Ali, A. and Ghahramani, A. |
| Journal Title | Journal of Environmental Informatics |
| Year | 2025 |
| ISSN | 1684-8799 |
| 1726-2135 | |
| Abstract | This study aimed to refine the estimation of Plant Available Water (PAW) by integrating satellite-derived data with limited soil moisture measurements in Millmeran, Queensland, Australia. The study focused on a paddock with a crop rotation of sorghum (summer) followed by barley (winter). MODIS 500m leaf area index (LAI) and the fraction of photosynthetically active radiation (fPAR) were used as inputs in two optional vegetation modules (Cover-based and LAI) of the HowLeaky soil water balance model. The model was calibrated and validated using MODIS 500m AET data and PAW computed from limited soil moisture measurements at 20 physical locations measured on two occasions. Six testing candidate models were built for every pixel within the paddock by combining two inputs (Cover and LAI) and three output options (AET only, AET+PAW, and PAW only). Sensitivity analysis indicated that fPAR and residue cover are crucial in the Cover module, while 14 parameters, including the radiation use efficiency (rue), are important in the LAI module. Models integrating both ET and PAW enhanced prediction accuracy. Specifically, the InCover-OutPAW [NSE (calib) = 0.848; AICc = 68] and InCover-OutPAWET [NSE (calib) = 0.903; AICc = 59] excelled in predicting PAW, with the InLAI-OutPAWET showing the best overall performance [NSE (calib) = 0.913; AICc = 54]. Using only MODIS AET for calibration and validation failed to predict PAW profiles effectively in cover module-based adaptation, whereas a satisfactory level of performance could be observed in LAI-based adaptation, subject to accurate characterisation of the LAI generic parameters. Integrating PAW as an observed variable significantly improved model accuracy, especially in capturing PAW variability post-rainfall. The study found a trade-off between precision and complexity, with models incorporating PAW measurements demonstrating improved prediction accuracy in a subhumid tropical setting of Australia. |
| Keywords | HowLeaky, Leaf Area Index, MODIS fPAR, Plant Available Water, PEST, Sensitivity Analysis, Soil Moisture, Australia |
| Contains Sensitive Content | Does not contain sensitive content |
| ANZSRC Field of Research 2020 | 300201. Agricultural hydrology |
| 370401. Computational modelling and simulation in earth sciences | |
| Byline Affiliations | Centre for Sustainable Agricultural Systems |
| Centre for Applied Climate Sciences | |
| Department of Environment and Science, Queensland |
https://research.usq.edu.au/item/100578/investigating-modis-derived-vegetation-metrics-on-the-howleaky-model-to-predict-plant-available-water-in-a-data-scarce-humid-subtropic-paddock-of-australia
0
total views0
total downloads0
views this month0
downloads this month