Repository for code accompanying paper Data Filling of Micrometeorological Variables in Complex Terrain for High-Resolution Nowcasting published in Atmosphere. Below is a description of the top level folders. Files with .m extensions are run with Matlab, and files with .ipynb extentions are run with Python 3 and Jupiter.
-
Averaging Period: Code studying the effect of the averaging period on performance. This was not published, but was requested by a reviewer. -
Combination Test: Calculate performance when using different input LEMS. Found in Appendix C. -
Combined Statistics: Code used to make all box plot figures in the manuscript. -
Correlation Analysis: Code used to make Figure A5 in Appendix D. -
Day Focus ANN: All ANN code is found here. There are separate scripts for each environmental variable, and separate plotting scripts for each environmental variable. -
Day Focus LinReg: All MLR code is found here. There are separate scripts for each environmental variable, and separate plotting scripts for each environmental variable. -
Day Focus RFR: All RFR code for Appendix E is found here. There are separate scripts for each environmental variable, and separate plotting scripts for each environmental variable. -
Hidden Nodes: Code used to analyze the number of hidden nodes required -
LEMS_Avg_Latest.mat: The main data file which can be read with Matlab. Only one copy of the data is included, and it may need to be copied/moved into run folders as necessary. The data is stored as a Matlab structure array. Code to read the data in can be seen in the various scripts throughout the repository. -
Location Sensitivity: Code to analyze the sensitivity to input LEMS locations as seen in Appendix C. -
Time Series Plots: Code to create time series Figures 5 and 6