• ${\color{red}T} im {\color{red}e}\ Se {\color{red}r} ies\ Extracti {\color{red}o} n\ for\ {\color{red}Po} lygonal\ {\color{red}Da} ta\ and\ Trend\ Analysis$

Name

• 🦖T(h)eroPoDa+ - Time Series Extraction for Polygonal Data and Trend Analysis ⬛

Description

• Toolkit created to extract median NDVI Time Series from Sentinel 2 data 🛰 stored in Google Earth Engine, perform gap filling and trend analysis

Author

• Vinícius Vieira Mesquita - vinicius.mesquita@ufg.br (Main Theropoda)

Co-author

• Leandro Leal Parente - leal.parente@gmail.com (Gap Filling and Trend Analysis implementation)

Version

• 1.1.0

Requirements (installation order from top to bottom)

• Python 3.10
• GDAL
• Rasterio
• Pandas
• Geopandas
• Scikit-learn
• Joblib
• Psutil
• scikit-map
• Earthengine-api

How to use

• In this version of TheroPoDa (1.1.0), you could extract a series of median NDVI from Sentinel 2 for a Feature Collection of polygons simplily by passing arguments to the python code exemplified below:

argument	usage	example
--asset	Choosed Earth Engine Vector Asset	users/vieiramesquita/LAPIG_FieldSamples/lapig_goias_fieldwork_2022_50m
--id_field	Vector column used as ID (use unique identifiers!)	ID_POINTS
--output_name	Output filename	LAPIG_Pasture_S2_NDVI_Monitoring_FieldWork

If you don't know how to upload your vector data in Earth Engine, you can follow the tutorial clicking this link.

Command line example

python main.py --asset users/vieiramesquita/LAPIG_FieldSamples/lapig_goias_fieldwork_2022_50m --id_field ID_POINTS --output_name LAPIG_Pasture_S2_NDVI_Monitoring_FieldWork

Roadmap

• Implement arguments to choose other zonal reducers (i.e. percentile, variance, etc.)
• Implement arguments to choose other satellite data series (i.e. Landsat series, MODIS products)
• Implement a visualization of the processed data (or samples of it)