SEDS at UIC

https://github.com/SEDS-UIC/nasa-space-apps-challenge-2022

https://github.com/SEDS-UIC/nasa-space-apps-challenge-2022

DASPA is a web application designed to give Mississippi river delta stakeholders an easy-to-use interface for socioeconomic prediction surface area maps. The application is comprised of HTML, CSS, JavaScript, MapBox, and a back-end API to our datasets and PyTorch-based general regression neural network model that outputs map layers onto the main application map. These map layers include expected land loss, expected land gain, mapped out expected population densities, expected population migrations, expected educational attainment within basin areas, expected median income within mapped out areas, and more. Users can input different values for certain ecological and socioeconomic factors using a series of sliders for the Atchafalaya and Terrebonne basins. Upon pressing the "Predict" button, DASPA takes the simulated conditions and runs it through our trained back-end model. This model's outputs are then piped to MapBox API and JavaScript functions to draw out surface area shapes on the main map. With a robust version of DASPA, we hope that policymakers can use the tool to view potential outcomes of area residents and create policy and legislation that can mitigate negative consequences of climate change.

We combined NASA's Delta-X airborne data campaign with datasets from NASA's Socioeconomic Data and Applications Center (SEDAC). The Mississippi river delta is losing land in some areas and gaining land in others. This is in no small part due to climate change consequences and we wanted to create something that could help the residents and policymakers of the Atchafalaya and Terrebonne Basins avoid disastrous consequences in the future. We used Delta-X readings including water level indicators, turbidity within islands, feldspar plots (soil accretion), Total Suspended Solids (TSS), and biomass datasets as inputs to a general regression neural network (GRNN) model along with median income, population density, educational attainment, unsatisfied basic needs, and median listing price datasets from SEDAC in order to generate area map layers with future average socioeconomic values. These map layers (such as expected land loss, expected land gain, expected population density, average educational attainment level within the basin area, and more) are then superimposed onto the web application's main map. We hope to make an impact in the lives that are directly affected by what is happening in the delta area with these prediction map layers.

We were inspired to tackle the airborne data challenge because our student organization is currently developing a cube satellite for remote sensing applications. We wanted to build a project that would teach us how to apply remote sensing data to a real-world project that can actually make a difference in the world. Our first brainstorming session was focused on not what we were going to make but how we were going to help people. This is how we came to the idea of DASPA, an application that takes NASA data and creates predictive models that can inform policymakers and researchers on how they can help constituents now. We divided the project into three main tasks: data investigation and cleaning, model development, and web application development. The datasets were large, the models demanding, and our student laptops were slow for the task. We split up data tasks to overcome our limited computing resources. We also had intermittent internet issues that made it more challenging to communicate as a team so we delegated tasks and checked in at regular intervals to keep development on track even if we went offline. We would like to thank Mudd Law Offices for being such gracious and generous hosts.

https://www.mapbox.com/

https://pytorch.org/

https://deltax.jpl.nasa.gov/

https://sedac.ciesin.columbia.edu/

https://css-tricks.com/snippets/css/a-guide-to-flexbox/

#deltax #sedac #socioeconomic #climatechange #earthscience #remotesensing