• The quality of the water can be measured with various methods and indicators, the concentration of chlorophyll in the water is one of them, one that is very related with human activity since the waste of urban areas usually contains many elements such as phosphorus, element that promotes the growth of phytoplantkon blooms that can harm both, the ecosystem and the people living nearby.
• These can be seen grow in populated areas, or areas with high human activity, such as tourist destinations.
• One of the most relevant touristic activities in México is visiting coasts, activity with a great influx of people in hotels those coasts.
• Due to the travel restrictions imposed because of the pandemic our hypothesize is that the water quality on Mexican coasts improves, and we can see that throughout the satellite data, specifically the maximum chlorophyll index.

*What did we develop? and what it does? and how it work?

With this idea in mind we developed a tool to collect satelital data from any available date range and masking only the relevant class of data, which in this case is the water area. We utilized API's from the ESA, specifically the sentinel's hub, but can be easily adapted to use other products, then the tool creates a series of images to generate a GIF to see the changes in the index, it also plots graphs of concentration changes over time.

One of the advantages of the tool is it's readiness to analyze any part of the world only changing the input coordinates of the desired area and the date range.

Why is this important?

Tools like this are important to see the effects of changes in human activities and how they are related with the change on our environment, and can lead to develop new studies and policies to improve and preservate the environment, which in this case could be policies for the hotel industry that impose new taxes to bad management of residual water discharged to the sea.

One thing is clear, if quality indicators show an improvement in the water with the restriction of human activity is because we have room of improvement.

What we hope to achieve?

We have to consider that the time of the event was limited, more in the specific case of our team, despite that we thing we made a prove of concept of a more and better integrated tool which can be included in the dashboard with a little bit more of work and love.

Main goal of the project has been to be able to see the changes over a timespan in water quality of economically important areas in order to generate solutions, if the indexes show a loss of quality due to human activities.

One future feature we'd like to include is automatic detection of zones where quality has dropped noticeably.

Team background story and inspiration

We had a rough start, first we knew about the challenge very little time before the event started, the I (Adrián) collect a group of friends in order to work in the water quality challenge since it is a common interests of us, then we set up a discord server to be in constant touch during the event, unfortunately i was let down by the other members who abandoned the project.

This was a problem since my background does not include much of programming, but i was convinced that i did't want to gave up, so i continued alone for various days trying to figure out the information provided by the tutorials and external coding tutorials and what is an API and how it works. Then at day 5, thanks to the chat channels i met Padminirai who joined late to the event and we decided to start this new team with a very clear objetive, analyze water quality in coastal areas of México.

It was difficult due to the shortened time we had compared with other teams, but our desire to participate and compete let us work hardly day and night for the rest of the days and we managed to build a working prove of concept.

What tools do we used?

We used the data provided by the Euro Data Cube, this includes Sentinel Hub API, the remote notebook, python 3.7, also some external libraries like proplot. In terms of hardware we worked with laptops, since having access to a remote computing unit of the EDC helped us a lot to compensate the lack of hardware power. In terms of software we worked everything within the EDC.

What problems and achievements we faced as a team?

The fist problem was previously described, but with the new formed team we faced the problem of being in different time zones, almost completely switched, this was difficult at the beggining, but we turned a disadvantage into an advantage, because with a little bit of better organization we managed to work continuously as it were shifts, when one start to finish his day, the other start to work on a new one, thanks to that, we achieved a good level of work and communication that helped us to compensate for our lack of experience with the subject and the lack of more members.

We used data collected from Sentinel's Hub API, specificaly, we used data from the satelite Sentinel 2A with the necessary bands for the water masking and the calculation of the Maximum Chlorophyll Index, those are B04, B05, B06, B11, SCL and CL.

We decided to keep using sentinel data because we had a little bit of experience using this products before, so we knew them better than others. The data influenced greatly our project, is the core of the analysis we made, but as i said previously, it can be easily switched to any other satelital product available throughout an API.

We tried to use sentinel 3 OLCI data, but the difference in resolution was too great for the spatial scale we worked on.

Slide: https://docs.google.com/presentation/d/1clzNEEtP3O35ywTxjugqLZN5wCD47maEX6R_Ei2RwD8/edit#slide=id.ge315890ebb_0_3

Tool code: https://github.com/adrian-drifter/EO_dashboard_hackathon_Second-chance-team.git

Our solution it's not ready to be integrated to the Dashboard, we aimed to build a prove of concept fo the tool, however, with a little bit more of work, we can further develop the tool until it's ready to be integrated, and add more capabilities to the same.

One feature that could be integrated are key zones where the client can open a pop up to select the dates he wishes to analyze the change of CHL of the time 1 and time 2 selected

Another option for future features is a machine learning integration to automatically detect when the indicators have a noticeable change.

sentinel hub: https://www.sentinel-hub.com/

EDC: https://eurodatacube.com/

#waterchallenge #api #sentinelhub #machinelearning #beaches #waterquality

This project has been submitted for consideration during the Judging process.