Mapping the Flames: Working with satellite data

by Joseph Frombaugh Posted on
Print 🖨 PDF 📄 eBook 📱

The Premise

In my GIS project on wildfires in Ontario and Quebec, one of the most valuable tools I have at my disposal is satellite data. Satellites offer valuable data that can be used for GIS projects such as this one, capturing data that allows us to assess the state of the climate, vegetation health, and many other things in near real-time.

Why Satellite Data Matters in Wildfire Research

Satellite imagery is indispensable for wildfire research. Here’s how it contributes to understanding wildfires:

  1. Mapping Burn Scars: Satellites like Sentinel-2 and Landsat capture high-resolution imagery that shows the areas affected by wildfires. By examining this data, we can map the spatial extent of burned areas, measure burn severity, and track changes in vegetation over time.
  2. Monitoring Vegetation Health: Healthy vegetation plays a major role in reducing fire risk, while dry or damaged vegetation can fuel fires. Satellite-derived vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), allow us to assess the health of vegetation across large areas. This data is especially valuable in identifying high-risk regions for future fires.
  3. Tracking Climate Data: Satellite data from projects like Copernicus can provide a clearer picture of climate patterns such as total precipitation, that could be relevant to understanding climate change and its ongoing role in exacerbating wildfires.
  4. Comparing Across Seasons: Since satellites have been capturing Earth imagery for decades, we can use historical data to compare past wildfire seasons with recent events. This allows us to identify trends and see how climate conditions are changing wildfire patterns over time.

Challenges in Accessing and Using Satellite Data

While satellite data is incredibly valuable, it can also be surprisingly difficult to obtain and use, especially when working with specific software like ArcGIS. Here are some of the challenges I encountered:

  1. Finding Free and Accessible Data: Many satellite datasets are open access, but finding the right dataset that meets your project’s needs can still be tricky. Data from sources like NASA’s FIRMS and the Copernicus Open Access Hub are freely available, but filtering through all available datasets to find one that covers the precise area, time frame, and type of analysis you need is time-consuming.
  2. Data Formats and Compatibility with ArcGIS: Even when you find the right data, it’s often not formatted in a way that’s easy to import into ArcGIS. Many satellite datasets come in formats like HDF, NetCDF, or GeoTIFF, which may require additional processing. Some formats work seamlessly with ArcGIS, but others require conversion through one or more of ArcGIS’s built-in Tools to ensure readability for import as a Layer.
  3. Processing and Storage Requirements: Satellite data files are often large, especially high-resolution imagery. Downloading, storing, and processing these files demands a lot of storage space and computing power. For example, a single scene from Sentinel-2 can be several gigabytes, and processing multiple scenes quickly adds up. Without powerful hardware or cloud resources, this can limit the scope of analysis.
  4. Data Pre-Processing: Before using satellite imagery for analysis, it often requires pre-processing, such as correcting for atmospheric interference, clipping to a specific region, or converting raw bands into indices like NDVI. While pre-processing ensures accurate results, it can add extra steps to the workflow, especially if you’re working with unfamiliar data formats.

Overcoming These Challenges

Despite these challenges, I was able to make some headway with satellite data by:

  • Relying on Open Data Platforms: Platforms like Copernicus Open Access Hub for Sentinel data and NASA Earthdata for MODIS and VIIRS data offer excellent resources, and with practice, the search and download process becomes easier.
  • Using ArcGIS Pro’s Data Interoperability Tools: ArcGIS Pro offers tools to convert data formats like NetCDF and HDF into more ArcGIS-compatible formats. This allowed me to bring in datasets that otherwise wouldn’t load properly.
  • Breaking Down Large Files: For storage and processing, I found it helpful to clip the data to focus only on specific study areas within Ontario and Quebec, which reduced file size and processing time.
  • Experimenting with Google Earth Engine: Google Earth Engine offers a cloud-based approach to accessing and processing large-scale satellite datasets. While it requires some coding skills, it’s an excellent option for efficiently working with satellite data, especially when you’re dealing with resource constraints.

Looking Ahead

Working with satellite data has its hurdles, but the insights it provides are invaluable. As I continue this project, I plan to explore more ways to integrate satellite data into my wildfire analysis, as well as other datasets I could use to further my research.

In the meantime, the above map I have created of precipitation in 2021, obtained from Copernicus, provides a glimpse into the kinds of satellite data I have been analyzing for this project and working with in ArcGIS.