Leveraging Google Earth Engine and Machine Learning to Estimate Evapotranspiration in a Commercial Forest Plantation DOI Creative Commons
Shaeden Gokool, Richard Kunz, Alistair Clulow

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(15), P. 2726 - 2726

Published: July 25, 2024

Estimation of actual evapotranspiration (ETa) based on reference (ETo) and the crop coefficient (Kc) remains one most widely used ETa estimation approaches. However, its application in non-agricultural natural environments has been limited, largely due to lack well-established Kc coefficients these environments. Alternate approaches have thus proposed such instances, with techniques use leaf area index (LAI) estimates being quite popular. In this study, we utilised satellite-derived LAI acquired through Google Earth Engine geospatial cloud computing platform machine learning quantify water a commercial forest plantation situated within eastern region South Africa. Various learning-based models were trained evaluated predict as function LAI, derived from best-performing model then conjunction situ measurements ETo estimate ETa. The ET comparisons against measurements. An ensemble showed best performance, yielding RMSE R2 values 0.05 0.68, respectively, when compared measured Kc. Comparisons between estimated yielded 0.51 mm d−1 0.90, respectively. These results promising further demonstrate potential provide robust efficient means handling large volumes data so that they can be optimally assist planning management decisions.

Language: Английский

Seasonal Drought Dynamics and the Time-Lag Effect in the MU Us Sandy Land (China) Under the Lens of Climate Change DOI Creative Commons
Fuqiang Wang, Ruiping Li, Sinan Wang

et al.

Land, Journal Year: 2024, Volume and Issue: 13(3), P. 307 - 307

Published: Feb. 29, 2024

Sand prevention and control are the main tasks of desertification control. The MU Us Sandy Land (MUSL), one China’s four deserts, frequently experiences droughts has a very fragile biological environment. Climate change is factor leading to drought, it may result in more serious drought situations future. Temperature Vegetation Dryness Index (TVDI) was established using land surface temperature normalized difference vegetation index data. In this paper, we investigate spatial temporal characteristics, future trends, time-lag effect TVDI on climate factors at different scales MUSL from 2001 2020 Sen + Mann–Kendall trend analysis, Hurstexponent, partial correlation lag analysis methods. results show that (1) overall shows characteristic gradually alleviating west east (TVDI = 0.6). A significant drying dominated 38.5% pixels fall (Z 1.99), highly rest three seasons average 2.95) whole year 3.47). (2) future, dry autumn, winter, will be by continuous drying, spring summer mainly wet. relationships between winter (−0.06) precipitation (−0.07) were negative, while evapotranspiration (0.18) showed positive correlation. six use types spring, summer, fall, primarily non-significantly positively correlated with evapotranspiration. (3) At seasonal scale, sensitive autumn opposite, responding quickly (0.3 months) being less (1.8 (2 months). interannual desert most (2.6 least responsive (3

Language: Английский

Citations

2
Leveraging Google Earth Engine and Machine Learning to Estimate Evapotranspiration in a Commercial Forest Plantation DOI Creative Commons
Shaeden Gokool, Richard Kunz, Alistair Clulow

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(15), P. 2726 - 2726

Published: July 25, 2024

Estimation of actual evapotranspiration (ETa) based on reference (ETo) and the crop coefficient (Kc) remains one most widely used ETa estimation approaches. However, its application in non-agricultural natural environments has been limited, largely due to lack well-established Kc coefficients these environments. Alternate approaches have thus proposed such instances, with techniques use leaf area index (LAI) estimates being quite popular. In this study, we utilised satellite-derived LAI acquired through Google Earth Engine geospatial cloud computing platform machine learning quantify water a commercial forest plantation situated within eastern region South Africa. Various learning-based models were trained evaluated predict as function LAI, derived from best-performing model then conjunction situ measurements ETo estimate ETa. The ET comparisons against measurements. An ensemble showed best performance, yielding RMSE R2 values 0.05 0.68, respectively, when compared measured Kc. Comparisons between estimated yielded 0.51 mm d−1 0.90, respectively. These results promising further demonstrate potential provide robust efficient means handling large volumes data so that they can be optimally assist planning management decisions.

Language: Английский

Citations

2