CASE STUDY
Machine Learning Platform for Redwood Prediction & Carbon Sequestration
An AI-powered environmental intelligence platform that predicts redwood distribution, estimates carbon sequestration, and visualizes ecological insights through interactive mapping.
CLIENT Delta Rising INDUSTRY Environmental Tech MY ROLE Machine Learning Data Scientist DURATION Nov 2024 – Feb 2025
96.9% Peak Model Accuracy
0.1629 RMSE
0.0309 MAE
18-Layer
CNN Architecture
15+
Environmental Variables
Interactive
GIS Platform
The Challenge
Coastal redwoods are among the world’s most important carbon-sequestering ecosystems. Accurately identifying where they grow and estimating carbon requires combining imagery, terrain, soil, and climate data.
THE GOAL
✓ Predict where redwood trees are likely to exist
✓ Estimate carbon sequestration potential
✓ Present insights through an interactive platform
Our Approach
We built a multi-source AI platform that integrates imagery, environmental layers, and forest carbon data through a dual-architecture neural network.
DATA
Collection
ENV
Layers
CNN
Prediction
CARBON
Model
ANALYTICS
Insights
INTERACTIVE
GIS
Model Performance
DEVELOPMENT STAGE R² SCORE RMSE MAE
Baseline CNN (Nov) 0.938 — —
+ Environmental Layers 0.969 0.1629 0.0309
+ Carbon Sequestration 0.967 0.0358 0.022
MODEL ACCURACY TREND ───────────────────────── 0.969
Key Outcomes
HIGH ACCURACY
Achieved up to 96.9% R² by integrating diverse environmental and carbon datasets.
COMPREHENSIVE PLATFORM
Combines computer vision, geospatial analysis, and ecological science.
ACTIONABLE INSIGHTS
Produces interpretable maps and metrics to support conservation decisions.
SCALABLE & REPRODUCIBLE
Built a repeatable pipeline that others can extend.
The 10-Step End-to-End Pipeline
01
Generate Points
02
Generate Images
03
Distance to Coast
04
Slope and Aspect
05
Soil & Land Cover
06
Climate Data
07
Probability Scores
08
Map Carbon Data
09
Model Analysis
10
Visualize Impact
Ready to Turn Data Into Impact?
Let’s build custom AI solutions that deliver real environmental impact.