9:00 AM - 10:30 AM
[ACG36-P02] Estimation of diurnal variation in terrestrial gross primary productivity across semi-arid ecosystems using Himawari-8
Keywords:Gross Primary Production, Himawari-8
Gross Primary Production (GPP) is the total amount of carbon dioxide absorbed by vegetation from the atmosphere through photosynthesis. Detailed estimation of GPP is important for understanding not only the terrestrial carbon cycle but also the environmental response of forests and agricultural land. Geostationary satellites can estimate surface biophysical status with high temporal resolution (e.g. 10 minutes). Therefore, geostationary satellite data can aid in estimating GPP on time scales of less than one day. However, the reproducibility of GPP under hot and dry conditions in arid regions remains a challenge. In this study, we attempted to estimate GPP in the Australian region covered by semi-arid vegetation using two GPP estimation models, the MODIS-GPP model and the EC-LUE model, with inputs of solar radiation from Himawari-8 and temperature and humidity data from a numerical climate model. The GPP estimation model was improved by evaluating its accuracy and optimizing the model parameters using GPP data observed at three OzFlux sites. The original model showed deviations from the observed values, but after optimization of the model parameters, the RMSE was reduced by 22% to 42%, confirming an improvement in accuracy. The EC-LUE model was able to better capture the characteristics of the site. However, it was difficult to adequately reproduce temperature and water stress measurements in semi-arid regions. This study indicates that the diurnal variation of GPP in semi-arid areas in the Australian region can be generally estimated. However, higher temporal resolution of meteorological data and improvement of the stress term in the model are needed for more precise estimation of environmental stress.