Time Series Extraction for Polygonal Data¶
Name¶
- T(h)eroPoDa + - Time Series Extraction for Polygonal Data and Trend Analysis ⬛
Description¶
- Toolkit created to extract Time Series information from Sentinel 2 🛰 data stored in Earth Engine, 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)