Changes between Version 16 and Version 17 of ImplementAnalysislnetf
- Timestamp:
- Jun 4, 2025, 11:16:32 AM (2 days ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
ImplementAnalysislnetf
v16 v17 44 44 == Overview == 45 45 46 The LNETF (localized nonlinear ensemble transform filter, Toedter and Ahrens, 2015) is a second-order exact particle filter with ensemble transofrmation (see, e.g. [wiki:PublicationsandPresentations Vetra-Carvalho et al., 2018]). There are different options to set perturbation noise or a stabilizing factor (`type_winf`, `limit_winf`) based on the effective sample size, see the [wiki:AvailableOptionsforInitPDAF#LNETFfiltertypePDAF_DA_LNETF10 Page on available options]. 47 46 48 For the analysis step of the LNETF algorithm, several operations related to the observations are needed. These operations are requested by PDAF by calling user-supplied routines. Intentionally, the operations are split into separate routines in order to keep the operations rather elementary as this procedure should simplify the implementation. The names of the required routines are specified in the call to the routine `PDAF_assimilate_lnetf` in the fully-parallel implementation (or `PDAF_put_state_lnetf` for the 'flexible' implementation) described below. With regard to the parallelization, all these routines (except `U_collect_state`) are executed by the filter processes (`filterpe=.true.`) only. 47 49