Hyperspectral Unmixing as an Analog Forecasting Method during Strong Monsoon Events in the Philippines
This repository contains all the codes and some data I used for my master's thesis on analog forecasting strong monsoon events in the Philippines using hyperspectral unmixing. The following is a short decription of what each folder contains:
This folder contains the Python codes applying analog forecasting and hyperspectral unmixing to a randomized training set for strong monsoon events from 2001 to 2018 in the Philippines. Training was applied for three domains: small, medium, large.
This folder contains the Python codes applying analog forecasting and hyperspectral unmixing to a randomized testing set for strong monsoon events from 2001 to 2018 in the Philippines. Testing was applied for three domains: small, medium, large.
This folder contains the Python codes applying a classic correlation analog forecasting on the small domain for both strong Amihan and Habagat. This was done to compare with our proposed method hyperspectral unmixing.
This is my full submitted manuscript.
I didn't include them here, but I obtained wind data from the NCEP reanalysis data provided by NOAA/OAR/ESRL PSL, Boulder, Colorado, USA.
Similarly, I obtained the mean daily SLP and RH from the JRA-55 reanalysis dataset. It can be downloaded from https://rda.ucar.edu/.
Lastly, the daily mean rainfall distribution was obtained from the GMP IMERGE which can be downloaded from https://disc.gsfc.nasa.gov/.
Just like my BS Thesis, I used the MATLAB code for hyperspectral unmixing provided in the paper:
J. Li, A. Agathos, D. Zaharie, J. M. Bioucas-Dias, A. Plaza, and X. Li. Minimum volume simplex analysis: A fast algorithm for linear hyperspectral unmixing. IEEE Transactions on Geoscience and Remote Sensing, 53(9):5067-5082, Sep. 2015.
Kindly refer to the following repository folder: https://github.com/cmdecastro/BSthesis/tree/main/MATLAB.
This is my full submitted manuscript.