trdougherty / tom.d

Machine Learning analysis of micro climate interaction with building energy consumption in New York City

TOM.D: Taking Advantage of Microclimate Data for Urban Building Energy Modeling

This github repo serves as a digital repository for the resources used in constructing the project: TOM.D. In this work, we explore the potential utility of open access data sources to aid in large scale energy modeling efforts. The process plow is as follows:

1. P0 - Data Collection
2. P1 - Data Cleaning
3. P2 - (Thermal) Collection of Microclimate Data - Google Earth Engine
4. P3 - Training Maching Learning
5. P4 - Model Interpretation
6. P5 - Research Questions

While following the steps in this work should automatically query the necessary data sets and deterministically rebuild the cleaned data, preprocessed data at each intermediate step has been provided in a Zenodo repository to ease the reconstruction effort.