Regional adaptation of a dynamic global vegetation model using a remote sensing data derived land cover map of Russia
The dynamic global vegetation model (DGVM) SEVER has been regionally adapted using a remote sensing data-derived land cover map in order to improve the reconstruction conformity of the distribution of vegetation functional types over Russia. The SEVER model was modified to address noticeable diverge...
| Published in: | Environmental Research Letters |
|---|---|
| Main Authors: | , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
IOP Publishing
2015
|
| Subjects: | |
| Online Access: | https://doi.org/10.1088/1748-9326/10/12/125007 https://doaj.org/article/be2bd03b29ca412b87cb7cd2c809cab8 |
| Summary: | The dynamic global vegetation model (DGVM) SEVER has been regionally adapted using a remote sensing data-derived land cover map in order to improve the reconstruction conformity of the distribution of vegetation functional types over Russia. The SEVER model was modified to address noticeable divergences between modelling results and the land cover map. The model modification included a light competition method elaboration and the introduction of a tundra class into the model. The rigorous optimisation of key model parameters was performed using a two-step procedure. First, an approximate global optimum was found using the efficient global optimisation (EGO) algorithm, and afterwards a local search in the vicinity of the approximate optimum was performed using the quasi-Newton algorithm BFGS. The regionally adapted model shows a significant improvement of the vegetation distribution reconstruction over Russia with better matching with the satellite-derived land cover map, which was confirmed by both a visual comparison and a formal conformity criterion. |
|---|