Tools for Analyzing Time Series of Satellite Imagery (TATSSI)

Install using Docker (recommended)

The Docker image ships the full TATSSI environment (Python 3.12, GDAL, PyQt, R + changepoint, earthaccess); the code and the data/ folder are shared with the host.

• Requirements:
  • Docker
  • xhost utils, to display the GUI through X11:
    • sudo apt install x11-xserver-utils
• Clone this repo
  • git clone https://github.com/tatssi/TATSSI
• Build the image (the first time takes a while):
  • cd TATSSI
  • docker build -t tatssi:latest .
• Run TATSSI:
  • ./run-tatssi.sh

Install using Anaconda (alternative)

You can install TATSSI on your favourite Linux distro or if you want to run it on Windows here you can follow the instructions to do it.

• Download and install conda if you do not have it already

• Clone this repo

  • git clone https://github.com/tatssi/TATSSI

• Create the environment from the environment.yml file:

  • cd TATSSI
  • conda env create -f environment.yml

• Activate the environment:

  • conda activate tatssi

• If you want to use the changepoint R package (used by the change point detection tools), install it in the environment R:

  • Rscript -e "install.packages('changepoint', repos='https://cloud.r-project.org')"

• Run TATSSI

  • If you want to use the Jupyter Notebooks:
    • Go to the TATSSI/notebooks directory and run jupyter notebook
  • If you prefer to use the UI:
    • Go to the TATSSI/TATSSI/UI directory and run python tatssi.py

Downloading products from LP DAAC with TATSSI

NEWS!

• LPDAAC went through a major update on July 31, 2023 as version 6.1 of several MODIS products were released. As of October 30, 2023 TATSSI provides direct access to these products.

Downloading products from the LP DAAC/Earthdata Cloud requires a NASA EarthData login. Please, first register as an EarthData user to get login credentials.

• If gedit is not installed in your system:

  • sudo apt install gedit

• Update config.json file with login credentials:

  • cd TATSSI/TATSSI/download
  • gedit config.json
  • Replace USERNAME and PASSWORD with login credentials, save and close

Description

TATSSI is a set of software tools to analise Earth Observation (EO) data. It allows you to:

• Download data from the Land Processes Distributed Active Archive Center (LP DAAC)
• New! Import GeoTIFF datasets (Landsat, Sentinel, RapidEye, WV2, etc. images)
• Transform to/from diverse EO raster data formats using GDAL
• Decode the QA-SDS associated to diverse MODIS & VIIRS data.
• Create time series of the aforementioned products masking by the user-defined QA parameter selection
• Perform basic gap-filling using the interpolation methods used in SciPy.
• Smooth time series using robust spline smoothing following Garcia. 2010
• Analyse time series using different tools such as decomposition, climatologies, trends, change point detection, etc.

There are some Jupyter Notebooks associated to each module, here you can find a description of each one.

Workshops materials

First workshop presentations (In Spanish)

• Introducción al manejo de calidad de datos
• Introducción a TATSSI
• Aplicaciones del análisis de series de tiempo
• Análisis de algunos métodos de interpolación

Second workshop videos (In Spanish)

• First day covered the Downloaders, Time Series Generation, QA Analytics, Interpolation and Smoothing TATSSI modules. YouTube video.
• Second day covered the Time Series Analysis TATSSI module. YouTube video.

Seminars & Presentations

• Red Latinoamericana de Teledetección e Incendios Forestales. YouTube video.
• 2020 Joint Statistical Meetings: Gerardo Lopez Saldana; Inder Tecuapetla.
• A quick glimpse of a simple plot for EVI and associated QAs.

Publications New!

• TATSSI: A Free and Open-Source Platform for Analyzing Earth Observation Products with Quality Data Assessment. International Journal of Geo-Information.
• Time Series Analyses for Sentinel-2 NDVI on Mangroves in Northwestern Mexico (2016-2022). IEEE 2023 Mexican International Conference on Computer Science.

Funding

TATSSI is funded by "Convocatoria de Proyectos de Desarrollo Científico para Atender Problemas Nacionales 2016" Project No. 2760; P.I.: Inder Tecuapetla. Collaborators: Gerardo Lopez Saldana, Rainer Ressl and Isabel Cruz.