The Amazon spans 2.1 million square miles of rain forest spread over nine countries. And on its edges are miles of agricultural fields, whose farmers routinely burn in order to control pests and weeds, and to encourage new growth.
Brazil is the largest cattle exporter in the world with over 200 million head of cattle. Ranchers often set fires to clear land for grazing, reports the Yale School of Forestry and Environmental Studies. However, these long-practiced techniques have raised concerns of the threat that accidental forest fires could pose during drought years. Climate change, which is beginning to show its effects around the world, could exacerbate this threat.
One way we can monitor and help to mitigate these dangers is by collecting detailed imagery and analyzing it. For many years, this was only possible on a rudimentary scale. But relatively recent advances in IoT, satellites, and geographical information systems (GIS) are allowing us to now track the fire location, intensity, and direction. This data collection begins with satellites in space, which literally take high definition pictures of the earth’s surface.
Sensors
MODIS (Moderate Resolution Imaging Spectroradiometer) is a spectroradiometer that measure both light amplitude and wavelength from sources such as fires. It has a wide swath which sees every point in the world every 1-2 days in 36 discrete spectral bands. MODIS records the frequency and distribution of cloud cover and aerosols in the atmosphere from forest fires. It can even distinguish flaming from smoldering burns. MODIS sensors aboard monitor several variables:
- Surface reflectance to provide an estimate of the surface spectral reflectance as it would be measured at ground level in the absence of atmospheric scattering or absorption. Low-level data are corrected for atmospheric gases and aerosols. The ‘Burned Area’ is calculated to identify burning areas on a per-pixel basis. An algorithm analyzes the daily surface reflectance dynamics to locate rapid changes and uses that information to detect the approximate date of burning, mapping the spatial extent of recent fires only.
- Land Surface Temperature (LST) and Emissivity are retrieved and estimated from land cover types, atmospheric column water vapor and lower boundary air surface temperature.
- Land Cover Dynamics product includes layers on the timing of vegetation growth, maturity, and senescence that mark the seasonal cycles.
Satellites
MODIS are mounted on NASA Terra and Aqua satellites. Terra satellites orbit around the Earth is timed so that it passes from north to south across the equator in the morning, while Aqua satellites pass south to north over the equator in the afternoon. NASA's Earth Observing System Data and Information System (EOSDIS) Worldview application provides the capability to interactively browse over 700 global, full-resolution satellite imagery layers and download the underlying data.
Mapping
Converting static raster images of fires into actionable intelligence requires GIS. Each image has both a time stamp in addition to location coordinates. It has to be cross-referenced with images from other satellites. This has to be done quickly in order for firefighters to respond effectively to the blazing fires.
When a commercial or government satellite captures this imagery, it is typically processed to a level that contains that raw data from the imagery and basic information about its location on the earth. This data is normally stored on large data libraries that exist in the cloud or at the provider’s location.
The GIS platform then imports this data and performs a series of steps to improve location accuracy, visual performance, and to prepare the data for interactive map presentation. These steps also include combining the data with elevation content to create a more accurate location that takes into account the characteristics of the surface and projecting that onto a two-dimensional plane appropriate for use within an interactive map.
GIS processes for importing data allow for multiple adjacent images to be imported and displayed together. This process is called “mosaicking,” and involves using the images corrected coordinates to place the imagery properly and correct for visual differences that can occur when images overlap. GIS handles the mosaicking process when the image is displayed and also supports the display of imagery that comes from multiple time periods. This “multi-temporal” data is the result of periodic refresh that occurs from government sources as well who have predicable rates of refresh depending on the satellite’s capabilities and mission. Commercial data providers refresh based on their own capabilities and the demands of their markets.
Interactive mapping applications within the GIS platform allow for multi-temporal data stacks to be visually accessed as well as used for change analysis. You can get a current view of the fires in the Amazon via a combination of satellite images processed with GIS:
ESRIIoT sensors on satellite collecting data on forest fires in the Amazon are part of the global nervous system to respond quickly to this disaster.