The COVID-19 pandemic offered an unprecedented opportunity to look at changes in the Earth system in response to reduced human activity. Your challenge is to develop tools to better understand changes in the interconnected Earth system as seen through the EO Dashboard.
The detection and surveillance of the consequences of human activity on water quality is essential for the care of resources, for this reason Matsú uses geospatial data and satellite images from the 3 space agencies (JAXA, ESA, NASA) to show the hidden side of COVID -19 from an overview. Matsú monitors the following three areas: (1-) Chlorophyll-a concentration of phytoplankton (Chl-a, mg / m3) from EO Dashboard. Feature: Geospatial Geometry Map. (2-) Oil spills from Sentinel Hub Acsess. Feature: Computer Vision Applied (3-) Water pollution caused by a turkey shipyard from Sentinel Hub Access. Feature: Multi-Image Gallery.
In this challenge we made the decision to cover different issues that are of great importance to improve the quality of water on our planet.
With the available resources of the challenge we propose a solution based on the monitoring of different factors such as chlorophyll using the water quality indicators and time series found in the Earth Observing Dashboard. Data from NASA (New York), ESA (Venice) and JAXA (Osaka) were used to determine anomalies in its concentration.
In addition, as part of the challenge, we are seeking information for monitoring fuel spills and pollution problems caused during the COVID-19 pandemic, such as the case of the Turkish shipyard.
Chlorophyll-a Concentration Anomaly Using Geospatial Geometry
The information in the EO Dashboard indicates raw data (as a .cvs file) and only a single maritime position (aoi). We consider that it is important to provide graphically different positions and their respective values in the polygon to be selected.
Computer Vision Applied on Water Pollution
This proposal is based on using Computer Vision techniques (OPEN CV) to detect oil spills in a satellite image, know the location of said disaster through a defined color palette and identify the size of the oil disaster caused in that place. To work on this solution we focused on the most recent spill (Mauritius Oil Spill).
Turkey Shipyard: Multi-Image Gallery
In western town of Aliaga, Turkey, we can see the increase in large amounts of scrap due to disuse of large luxury cruise ships. Billions of dollars' worth of luxury vessels are dismantled, but the waste/spillages are going directly into the Mediterranean Sea which is already under eco-strain. What we can do? It is important to record the future implications of this situation where ecological damage is evident in the area.
We are a team of four young women from high school and our tutor, who since 2020, and during the pandemic caused by COVID-19, began training in the use of images with open source computer vision projects and manipulation tools and analysis of data. We believe in a fast, powerful, flexible and easy to use, built on Python programming language.
Our inspiration began with developing more dynamic alternatives and using them with images and data. Since 2019 we have also worked with boards to share the response of sensors in integrated systems of the Internet of Things.
The development of improvements for the EO Dashboard is to propose the use of images for the visual monitoring of certain anomalies, contamination and detection of possible future problems of the effects on the mortality of Zooplankton, hydrocarbon content due to human industry, and of course continue with the chlorophyll content.
During the challenge process, we were in a totally remote communication due to the policies and mobility restrictions in the face of the pandemic in our country. Each of us worked with her own team, updated dependencies locally, and tested the code on the EOX server.
We are very happy that the challenge allowed the use of Python libraries. We recognize that using Jupyter Lab with fixed dependencies was new to us. For example, during the duration of the challenge, we learned about Cartopy instead of Basemap, which was not available.
We are grateful that the opencv library was there to be able, even in a small code, to show our learning and work.
We consider ourselves with the drive and dedication to develop information boards.
The challenge is a learning process, and therefore, we not only focus on using the data that NASA offers in the EO Dashboard about water quality indicators.
Our work included the investigation of all available resources that refer to the data shown (even those provided by media) The data were manipulated with a pandas library for analysis and using matplotlib to plot them over time.
It was detected that NASA records are daily, and therefore, it is possible to create a large database (Big Data) that can, in its future stage of development, apply deep learning with them.
We accessed geospatial data provided by the Japanese agency JAXA and learned how to manipulate .nc files. Although we do not present them in the final code, we consider that the cryosphere is another element that could be included in the EO Dashboard.
For the use of true color satellite images and what bands were needed, we used the Documentation of Sentinel Hub Python package for ESA satellites.
Video link:
Our proposal is to integrate three possible uses of geospatial data:
Code 1) Water Quality Time Series Chlorophyll-a concentration anomaly creating closed spatial polygons by the user for to view points of interest.
Code 2) Apply computer vision techniques to observe the impact of economic disasters with machine learning.
Code 3) Create a binnacle of images of possible human activities that are affecting the environment, even when these were as a result of covid-19.
Code Submission link:
https://drive.google.com/drive/folders/1WTaRZbQ4ik7PT_8pMCoZMofzmDazn1xo?usp=sharing
Water Quality Indicators data files from Earth Observing Dashboard: https://eodashboard.org/
TE-Global: https://www.eorc.jaxa.jp/water/index.html
NASA Earth Data: https://earthdata.nasa.gov/
Images Analysis: https://makersportal.com/blog/2019/7/8/satellite-imagery-analysis-in-python-part-i-goes-16-data-netcdf-files-and-the-basemap-toolkit
Using Cartopy: https://towardsdatascience.com/plotting-geospatial-data-with-cartopy-4b5ad0da0761
Understanding Ocean Images: https://oceanwatch.pifsc.noaa.gov/erddap/griddap/noaa_snpp_chla_daily.html
Request from SentinelHub: https://sentinelhub-py.readthedocs.io/en/latest/examples/process_request.html
OpenCV Algorithm: https://opencv.org/
Matplotlib: https://matplotlib.org/stable/tutorials/introductory/images.html#sphx-glr-tutorials-introductory-images-py
Satellite Imagery in Python: https://github.com/acgeospatial/Satellite_Imagery_Python
Chlorophyll Concentration: https://neo.sci.gsfc.nasa.gov/view.php?datasetId=MY1DMM_CHLORA
Tutorial for use NetCDF: http://blog.cityprojections.com/2017/07/tutorial-for-working-with-netcdf.html
Scientific Visualization: https://svs.gsfc.nasa.gov/4820
NASA FORUM: https://forum.earthdata.nasa.gov/
#human_activity #water_quality #chlorophyll #satellites #pollution #environmental_disaster
This project has been submitted for consideration during the Judging process.