About FRM4FIRE

Vital information for emergency services

In the context of fire monitoring using remote sensing satellites, burned area (BA), active fire (AF), and the associated Fire Radiative Power (FRP) products can be retrieved to characterise fire activity. Each of these products relies on different sensor data and processing algorithm approaches.

While BA products are binary classifications flagging a temporal land surface change (burned scars), AF products are based on threshold approaches that detect hotspot anomalies. FRP is a derivative of AF, quantifying the rate of radiative energy released due to a fire. Whereas BA is aimed at ensuring area completeness, AF and FRP provide an instantaneous location and ‘intensity’ of an ongoing fire, providing emergency services vital information, especially useful during night-time.

The Sentinel-3 SLSTR Fire Radiative product

Sentinel-3 is a European Earth Observation satellite mission that supports Copernicus ocean, land, atmospheric, emergency, security, and cryospheric applications. The mission is jointly operated by ESA and EUMETSAT to deliver operational ocean and land observation services. Among several objectives, ESA Sentinel-3 provides global coverage of fire monitoring products, such as FRP. Characterised by large swath coverage and high overpass revisit, the mission is made up of a constellation of two identical satellites, Sentinel-3A (S3A, launched in 2016) and Sentinel-3B (launched in 2018).

The Sea and Land Surface Temperature Radiometer (SLSTR) sensor onboard Sentinel-3 has two dedicated ‘fire channels’, that combined with the other channels, detect and characterise active ongoing fires, for which Near Real Time (NRT) and Non Time Critical (NTC) are available, processed by EUMETSAT and ESA, respectively. This provides users a comprehensive set of fire detections expected to detect fires within about 10^-4 of a pixel within an image. This equates to an FRP of around 8 MW and an active fire area of about 100 m².

The accuracy of Sentinel-3 SLSTR based Fire Radiative Product (FRP) data is directly influenced by the quality of on-board fire detection instrument calibrations. The data produced contains full calculations and delivery of associated uncertainties, but the true 'quality' of the product in terms of ability to characterise areas of fire activity and fire emissions to the atmosphere remain unknowns.

Instrument Calibration and Validation

CEOS defines Calibration as, ‘the process of quantitatively defining a system’s responses to known, controlled signal inputs’. Validation is the process of assessing, by independent means, the uncertainty of data products derived from those system outputs — so is a core component of satellite missions.

Without adequate validations, geophysical retrieval methods, algorithms, and geophysical parameters derived from satellite measurements cannot be used with confidence. As such, returns on investment for the satellite mission are limited. In addition, meaningful uncertainty estimates cannot be provided.

In orbit, uncertainty characteristics of the satellite instruments (i.e., imaging radiometers, GPS, etc.) previously established — by pre-launch laboratory calibrations, characterisations, and end-to-end instrument measurements (e.g., brightness temperature) from which geophysical measurements (e.g., Fire Radiative Power) are retrieved — may only be assessed via independent calibration and validation activities.

Validation exercises for FRP products normally provide comparisons with other FRP products, either statistically or per fire-event, by combining geostationary and polar orbiting derived data to restrict comparisons to near simultaneous detections. In addition, such comparisons may only be conducted on coincident unsaturated radiance measurements.

Under the current CEOS Land Product Validation (LPV) framework, FRP products are at stage 1 validation, and it is not expected that FRP products would achieve stage 3 in the near future. Due to the short-lived nature of fires, generating ground base reference data is the major challenge, making UAVs and airborne thermal imaging systems (such as the FIDEX experiment) the preferred option. However, because of the resources required, these exercises are rare and in only in initial stages of development, lacking suitable measurement protocols that may ensure best practice guides are followed, so that data may be fully traceable, uncertainties are properly budgeted, and exercises may take place under the best possible conditions.

There's a clear need for Fiducial Reference Measurements for FRP.