search

zhenyicun / 3DA_code

This is the code for the paper "An adaptive optimal interpolation based on analog forecasting: application to SSH in the Gulf of Mexico"
https://zenodo.org/record/3559784#.Xk0BIeko85l
3 stars 2 forks source link

readme

🌍️DESCRIPTION OF FOLDERS: /data/OCCIPUT/ : the extracted 50-member OCCIPUT simulation at GoM /data/obs/ : the extracted real tracks position and time of the altimeters in 2004. It also saves the truth and the SSH along-tracks after you run get_obs.py /data/catalog/ : is used to save the catalog for AnDA /data/results/AnDA/ : is used to save the results of the last run of AnDA /data/results/OI/ : is used to save the results of the last run of OI /data/results/OI_COA/ : is used to save the results of the last run of OI_COA (following the OI described in <LeTraon et.al. 1998> ) /figs/ : saves all the diagnostic figs.

🌍️DESCRIPTION OF EXISTING RESULTS: /data/results/AnDA_k1000_ROI10/ : the result of AnDA where k=1000, and spatial localization radius = 10 degress /data/results/OI_6/ : the result of OI where temporal correlation scale = 6 days /data/results/OI_10/ : the result of OI where temporal correlation scale = 10 days /data/results/OI_15/ : the result of OI where temporal correlation scale = 15 days

🌍️DESCRIPTION OF PARAMETERS:

🌐️AnDA: max_mode : number of EOFs used in the state variables do_KS : "yes"---do Kalman smoother, "no"---only Kalman filter Ne : ensemble size for EnKF and EnKS R : observation error variance for EnKF do_localization : "yes"---do covariance spatial localization for EnKF and EnKS, "no"---no localization
rloc : localization radius for covariance spatial localization used in EnKF and EnKS AF_k : number of nearest neighbors to be search for in the Analog forecast algorithm

🌐️OI: r_spat
s_spat r_temp s_temp : temporal correlation scale R : observation error variance used in the algorithm

🌍️RUN THE CODE:

1, run get_obs.py: generate the truth using OCCIPUT member#1 at year 20, and the ssh at the given tracks at the given time

2, run get_catalog.py: generate the catalog using OCCIPUT members#2-50, from year 1 to 19

3.1, choose the parameters of AnDA, OI, and OI_COA by editting main_AnDA.py, main_OI.py, and main_OI_COA.py 3.2, run main_AnDA.py, or main_OI.py, or main_OI_COA.py. The new results will be saved in /data/results/AnDA/, /data/results/OI/, and /data/results/OI_COA/

4, run generate_figs.py to generate all the figs and check the RMSE. You will see all the figs in this paper except the two spectrum figs produced by Sammy.