Project methods overview

Data Sources

• CBI (Composite Burn Index)
• LCMAP (Land Change Monitoring, Assessment, and Projection): Land cover information
• CU Boulder Disturbance Stack
• CHELSA V1.2: climate data
• NASADEM from Microsoft Planetary Computer for topographic data
• GEDI for biomass data

Data Processing Steps

Describe data processing steps taken, the order of scripts, etc.

Data Analysis

We will take our processed data and fit an XGBoost model with it (initially with default parameters), to get measures of performance. Then, using the shap Python package, we plot the shap values to visualize each feature's impact on the model output. If this model does well, we can then use it to predict the land cover change and resulting carbon loss of arbitrary future wildfire events.

Visualizations

Describe visualizations created and any specialized techniques or libraries that users should be aware of.

Conclusions

Summary of the full workflow and its outcomes. Reflect on the methods used.

References

Citations of tools, data sources, and other references used.

---

Last update: 2024-05-09