11月1日 | Xiaogu Zheng：Global Carbon Assimilation System using a local ensemble alman filter with multiple ecosystem models

时   间：2023年11月1日 10:30-11:30

地   点: 普陀校区理科大楼A1514

报告人：Xiaogu Zheng 上海张江数学研究院教授

主持人：项冬冬  华东师范大学教授

摘   要：

Dr Xiaogu Zheng was a senior scientist at the National Institute of Water and Atmospheric Research NZ and a professor and chief scientist at the College of Global Change and Earth Systems Science Beijing Normal University. He received a PhD in statistics at Beijing Normal University and was a postdoctoral fellow at Victoria University of Wellington NZ. He is the recipient of 2007 Edward Kidson Medal awarded by NZ meteorological Society. He published about 100 papers in meteorology and 40 papers in statistics.

<img nvolved="" in="" developing="" madden’s="" potential="" predictability="" theory="" and="" its="" applications="" to="" s2s="" prediction="" multidecadal="" peak="" detection;="" estimating="" forecast="" error="" statistics="" for="" data="" assimilation;="" mapping="" temperature,="" rainfall,="" frost="" risk="" wind="" energy;="" introducing="" climate="" variability="" weather="" generators;="" trend="" analysis.<="" span="">

报告人简介：

In this study, several ideas for improving the process of carbon fixation by photosynthesis are studied, such as including atmospheric CO₂ concentration in state vectors, using one week assimilation window in the Ensemble Kalman filter (EnKF), using analysis states to iteratively estimate ensemble forecast errors, and a maximum likelihood estimation for the inflation factors of the forecast and observation errors. These ideas are tested using an assimilation system similar to CarbonTracker，but with different ecosystem and atmospheric transport models. It is found that these ideas are reasonable and potentially useful. Based on the frame work of CarbonTracker combined with these innovative ideas, a Global Carbon Assimilation System using a modified Ensemble Kalman filter (GCAS-EK) is proposed. GCAS-EK is then used to estimate the terrestrial ecosystem carbon fluxes from 2002 to 2008. The results showed that this novel approach can effectively reduce the biases and uncertainties of the carbon fluxes simulated by the ecosystem model.