Problem: Current deforestation efforts are random and excessive. This means that it is worse and unnecessarily harmful to the environment. Companies will deforest no matter what for the next several decades. Efforts need to go towards smart deforestation.
Existing solutions:
Our solution: Our model takes in deforestation data and biomarker data to monitor current deforestation, predict effects of future deforestation, and make optimized choices on where governments and private organizations should implement required deforestation when necessary.
Monitor environmental change: RUBRIC: Use Earth Engine’s satellite datasets to detect and visualize deforestation, urban expansion, or land degradation in near real time. Gemini how we achieve this point.
Predict and prepare: RUBRIC: Combine Earth Engine’s geospatial insights with AI or machine learning to forecast extreme weather events, droughts, or flood risks — enabling earlier, data-driven responses. Gemini how we achieve this point.
Optimize for sustainability: RUBRIC: Integrate geospatial data to improve systems like agriculture, logistics, or energy distribution (e.g., smarter food delivery routes or resource allocation that minimizes emissions). Gemini how we achieve this point.
How this works: We are taking two dataset pairs and training the model on 2017 data for NDVI + EVI with MOD44B then comparing it to the 2020 data for the NDVI + EVI with MOD44B. The model will learn to compare the surrounding regions of NDVI + EVI data depending on which parts of the forest got deforested Can use this to determine which parts of the forest are safest to deforest in the case of necessary deforestation
Greater Vision: Deforestation is inherently bad for the environment. It kills surrounding ecosystems. Current deforestation is horrible, it displaces humans and animals alike, lays ruin to fragile biomes, and has enormous downstream effects. There is a better way. For the foreseeable few decades we will need to be deforesting to maintain a lot of the supply chain. We need to invest efforts into optimizing deforestation to reduce effects. The Dept. of Forest Efficiency (DOFE) will work with private orgs and companies to help them deforest better until we can wean ourselves from this malignant practice.
Q and A: What is necessary for deforestation? Will we always need to deforest?
Technicals: Data for our model is The forest cover change between 2020 and 2017 from our google dataset (resolution of 250km) The NDVI change in the surrounding areas (for a certain radius of km) NOT of the specific deforested region Maybe same for EVI
Built With
- gemini
- googlearthengine
- javascript
Log in or sign up for Devpost to join the conversation.