Programming

Montréal, Canada

Automating individual tree detection in parks using LiDAR data and open source tools

Urban parks and forests provide a string of ecosystem services to city-dwellers from reducing the effect of the urban heat island to being a proven reliever of mental stresses. Mapping the inventory of urban trees within a city is incredibly helpful to keep these spaces accounted for and healthy.

Check out this R script on GitHub to learn how to:

  • access and download LiDAR data for the City of Montréal

  • use the lidR package to create a LAS file catalogue and visualize point clouds

  • clip point clouds to the extent of park boundary sf objects collected using the osmdata package

  • subset points by their classification

  • detect tree tops and visualize locations in an interactive leaflet map

Thimphu, Bhutan

Automating haze removal, atmospheric correction and mosaicking of satellite imagery in Python

Using mainly the PCI Geomatica Python library, any amount of raw, Level-1 Landsat 8/9 images can be fed into this script and the images will be mosaicked into a clean, colourized composite image of a larger region, in this case, a view of Bhutan.


New York, NY

Using geospatial Python libraries to examine NYC subway vulnerabilities to climate change

This final project for a programming class used several python libraries for the processing, analysis and visualization of data relating to subway stations entrances (underground, elevated), neighborhood tabulation areas (NTA), and past & projected floodplain data for the city of New York. Libraries used include:

  • Pandas

  • GeoPandas

  • fiona

  • shapely

  • numpy

  • bokeh

  • mplleaflet

  • folium

The python scripts also provide functions to automate the creation of consistently styled maps using user-input data.

Please consult the report for more information on the methods used and results.