A tutorial session was offered in the neXtSIM workshop in 2025. In this tutorial, we show how DA works for neXtSIM-DG in an Observing System Simulation Experiment (OSSE):
NEDAS provides a light-weight Python solution to the ensemble data assimilation (DA) problem for geophysical models. It serves as a new test environment for DA researchers. Thanks to its modular and scalable design, new DA algorithms can be rapidly tested and developed even for large-dimensional models. NEDAS offers a collection of state-of-the-art DA algorithms, including both serial and batch assimilation approaches. NEDAS offers DA researchers with new ideas to test their prototypes early-on in real-model-like environments, before committing resources to full implementation in operational systems.
parallel_python_tutorial is a short demonstration of how to use Python libraries to achieve parallelism and speed up a program: multiprocessing, threading, concurrent.futures, mpi4py, and dask.
In NERSC "Crash Course on DA", I gave this specific tutorial on multiscale data assimilation.
A two-dimensional vortex with simple vorticity dynamics, for testing multiscale data assimilation algorithms with alignment techniques: rankine.
A multiscale DA algorithms testbed written in Python, the QG model from Dr. Shafer Smith is used as the dynamical system: QG_Multiscale_DA
The Pennsylvania State University Ensemble Kalman Filter system for the Weather Research and Forecasting model (PSU_WRF_EnKF) originated from Fuqing Zhang's project code when he was a postdoc at NCAR. The project was a proof of concept for radar data assimilation for the convective-scale weather. It quickly evolved and became a reseach testbed for ensemble data assimilation methods and their application to various weather systems. It was also developed into a quasi-operational hurricane ensemble forecast system. The PennState ADAPT center currently host the PSU_WRF_EnKF system development team, and aims to further apply data assimilation and predictability techniques to many fields beyond weather prediction.
The PSU_WRF_EnKF code should be downloaded from the official Code Release Page.
I contributed to the development of the PSU_WRF_EnKF system as Fuqing's PhD student. In 2012, I rewrote the Fortran MPI parallel algorithms and increased the efficiency of the system. I also participated in the development of assimilation capability for satellite radiance data and many other types of observing systems. By the time I graduate in 2018, I wrote this Technical Note to comprehensively document the current system and describe its components.
A toy system for testing data assimilation algorithms using the Lorenz-96 model: L96_DA
Matlab code for demonstration of the shallow-water systems: 1D model, 2D model
In 2009, as I started my master degree program in Peking University, I learned how to use high performance supercomputers and worked as a part-time system administrator for a small 32-node cluster hosted in the Department of Atmospheric Science. I wrote this
Cluster User Guide V2 (in Chinese) to help fellow students learn how to use the cluster system for their research. The document received more attention than I expected, it was even circulating among students and researchers outside Peking University.