Data and Code for Bayesian Auto-regressive Distributed lag models

Data and Code repository for work on a paper titled: Forecasting Vegetation Condition with a Bayesian Auto-regressive Distributed Lags (BARDL) Model.

Each data file has a time series (12-week moving average) of data extracted from the satellite earth observation datasets over some selected counties in Kenya. mages were corrected and cleaned by removing all clouded pixels. Gaps have been filled and Smoothed.

Each file has the following columns:

SoilMoist: Daily Soil Moisture data from ESA CCI combined soil moisture products
LST: Weekly (3 Month rolling average ) Land surface temperature from MODIS, MOD11A1.v06, products
Rainfall Estimates: Weekly (3 Month rolling average ) Rainfall estimates from the CHIRPS repository
NDVI: Weekly (3 Month rolling average ) Normalised Difference Vegetation Index from MODIS MCD43A4.v60, products
SAVI: Weekly (3 Month rolling average ) Soil Adjusted Vegetation Index from MODIS MCD43A4.v60, products
VCI: Weekly (3 Month rolling average ) Vegetation Condition Index from MODIS MCD43A4.v60, products
TCI: Weekly (3 Month rolling average )Temperature Condition Index from MODIS MCD43A4.v60, products
VHI: Weekly (3 Month rolling average ) Vegetation Health Index from MODIS MCD43A4.v60, products

All columns with *_Anom are the anomaly version of the time series.

All datasets were prepared by @edd3x (Email: e.salakpi@sussex.ac.uk)