A new regional reference evapotranspiration model based on quantile approximation of meteorological variables
Reference evapotranspiration (ETo) is a variable that can assist in estimating agricultural water use in water-scarce regions. Estimating ETo with limited data is an important alternative to overcome the current shortage of meteorological data in many areas around the world. For this purpose, this s...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Agricultural Water Management |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0378377425000137 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Reference evapotranspiration (ETo) is a variable that can assist in estimating agricultural water use in water-scarce regions. Estimating ETo with limited data is an important alternative to overcome the current shortage of meteorological data in many areas around the world. For this purpose, this study introduces a new method for establishing a simplified regional ETo model. The method, which creating ETo models based on temperature at meteorological stations that have the highest quantile matching with the target station's meteorological variables based on the closest meteorological data characteristics. To test the performance of the new method, we used data from 120 meteorological stations in Northwest China from 2000 to 2021 to develop XGBoost models to establish the new regional ETo model. We compared the proposed method with local models and two conventional regional ETo models to evaluate its performance. While the new method increased the Root Mean Square Error (RMSE) by an average of 13.4 % compared to local models, it demonstrated significant advantages over conventional regional models. Specifically, the RMSE decreased by 6.4–7.1 %, the Normalized RMSE (NRMSE) decreased by 5.5–7.3 %, computation time was reduced by 18.4–21.8 times, and spatial memory usage was reduced by 147–211 %. These improvements make the proposed method more efficient and scalable, particularly for regional applications in data-scarce areas. |
|---|---|
| ISSN: | 1873-2283 |