DA_testbed for quick prototyping

How to use

This repository should only have core features. Each experiments should be conducted after folking/branching this repository.

make
evince image/merged.pdf

main: run nature run, observation, and DA cycle. Output is saved at data/
da_system: wrapper for DA algorithms. It can deal 3D/4D-DA and deterministic/ensemble DA.
- letkf: LETKF core programs
- ensrf: Sequential EnSRF core programs
- fdvar: 4D-Var core programs
- tdvar: 3D-Var core programs
model: model integration and TLM
obs: observation generation and xb -> yb transform.
- class Scaler_obs allows flexible observation networks, localization, and 4D-implementation
plot: all visualization functions. Read data/ and write image/
const: Run-time settings

Single-process
- For parallel execution of multiple experiments, it should be externally wrapped with https://github.com/ysdtkm/all-in-one/tree/master/works/2018/parallel_wrapper
Numpy, Numba for acceleration. No use of ctypes/f2py