Skip to content
All challenges

Harmonization of MODIS and VIIRS Hot Spots

IntermediateAdvancedSubjects:Earth ScienceSoftware

Your challenge is to build a web application that harmonizes these records into a burning activity calendar based on active fire hotspots, enabling early warning and emergency responders, scientists, and land managers to examine historical fire patterns, unusual conditions, and critical periods within a selected area of interest.

15 - Harmonization of MODIS and VIIRS Hot Spots_4.3
Image: The National Commission on Space Activities of Argentina (CONAE)

Summary

Understanding how burning activity is distributed across our planet over time, including wildfires and agricultural burns intense enough to be detected by satellite sensors, enables us to anticipate critical periods, guide monitoring efforts, and improve fire response. Satellites have tracked active fire hotspots for more than two decades, but the record is split across sensors whose data cannot be compared directly, leaving fire managers without a consistent picture of when and where burning has occurred. Your challenge is to build a web application that harmonizes these records into a burning activity calendar based on active fire hotspots, enabling early warning and emergency responders, scientists, and land managers to examine historical fire patterns, unusual conditions, and critical periods within a selected area of interest.

Plays from YouTube

Background

Fire regimes

Some ecosystems need periodic fires to maintain their function and their structure. However, changes in fire intensity, extent, and frequency can have negative impacts on climate, society, and biodiversity, along with the ecosystem services they provide. Fires can also cause loss of life and properties, degradation of health, and acute risk to fire managers.

Because humans have modified global patterns of vegetation, ignition, and climate over millennia, fires on both local and global scales are increasingly human-related.

Burning activity involves wildfires in natural vegetation and forested areas, as well as anthropogenic fires associated with agriculture, land-clearing, waste disposal, and industrial activities. It is one of the many dynamic phenomena that can be detected with remote sensing.

Satellite fire detection and monitoring

Climate models project that fire weather will become more common and more intense. Earth Observation (EO) satellites provide technological solutions for early detection, rapid response, and postfire assessment, measuring thermal radiation and smoke across regional to global scales.

Detecting hotspots on Earth's surface

EO satellites provide information on fire location, intensity, and temporal evolution. Satellite sensors operating in the thermal infrared (TIR) range can detect hotspots on the Earth's surface generated by urban and wildfires, volcanic eruptions, and industrial processes such as gas flares in oil and gas operations.

For active fires, detection and monitoring rely mainly on mid-infrared (MIR) and TIR measurements, where active combustion produces its stronger signal. Sensors such as MODIS and VIIRS provide real-time thermal anomaly detection, enabling early hotspot identification and the assessment of burned areas.

Burn activity calendars

Using satellite-detected hotspots, fire activity calendars can be developed by integrating historical information with recent data, allowing for the rapid visual identification of seasonal patterns, anomalies, and trends in each region.

MODIS sensors on board Terra and Aqua satellites provide more than two decades of observations, but they are approaching the end of their operational life and are being replaced by VIIRS sensors. Their records cannot be directly compared without harmonization, since MODIS and VIIRS have different spatial resolutions, sensitivities, and observation schedules.

Therein lies the root of the challenge. As fire frequency and intensity increase, emergency managers benefit from an updated and consistent hotspot calendar that enables them to rapidly visualize, for example, which month and years had the highest fire activity, and compare the burning activity of the current month against the historical median for a selected area.

You may consider harmonizing NASA Earthdata FIRMS hotspots, combining the extensive historical series of MODIS with the continuity and greater detail of VIIRS.

Emergency managers could then use the web application to monitor sensitive regions, recognize changes relative to historical patterns, and strengthen fire forecasting and response.

Objectives

Your challenge is to develop a harmonized burning activity calendar based on active fire hotspots in a web application, combining the extensive historical series of MODIS with the continuity and greater detail of VIIRS to provide a consistent time series of hotspot activity.

This challenge could be approached in two parts. The first is MODIS and VIIRS hotspot harmonization, so that by the end of that phase the sensor of origin no longer matters and the patterns, trends, and anomalies become the focus.

The second involves developing a web application that takes harmonized hotspot data as input. Through this visual tool, a user could select a region and a specific date range, then download analytics and visualizations of fire activity.

A calendar of burning activity showing the number of hotspots detected from the early 2000s to the present, built on the harmonized hotspots dataset and integrated into the web application, would give emergency managers a view of the fire history of an area: which periods of the year see more fires, which years had more burning activity, and how a current fire compares with previous ones.

In addition to numerical information, you could consider how color ramps might represent similarities and differences in the number of detected hotspots across months, providing an accessible and intuitive way to visualize fire activity. A downloadable CSV File of hot spots detected during a search would be another useful output.

Potential considerations

You may (but are not required to) consider the following:

  1. Your target audience can include early warning and emergency responders, scientists, land managers, and firefighters.
  2. Consider making your web application interactive – for example, a map where users can draw the area of interest, or an option to enter latitude and longitude.
  3. Consider using accessible color ramps.
  4. Your tool could depict a burning calendar
  5. Your tool could focus on a few key indices and figures and present them through data visualizations, such as:
    • Historical comparisons, including anomalies against multi-year averages
    • Number of fires detected and reported
See the 1 resource for this challenge

From the NASA Space Apps Challenge. The official page is the final word on this challenge.

Built in Athlone

The 3 projects Athlone teams submitted for this challenge at HackAthlone 2026, from the teams’ own Space Apps pages. Links open in a new tab.

    • 2026

    DrySpell: Harmonized MODIS + VIIRS Fire Calendar

    by DrySpell

    Harmonization of MODIS and VIIRS Hot Spots

    DrySpell is a web app that turns two generations of NASA satellite fire detections into one consistent fire-activity calendar. MODIS has recorded active fires since 2000; VIIRS is newer and much more sensitive, so simply joining the two records creates a fake jump in activity. We put both on a common 0.25° grid, count detected cell-days per month, and fit one VIIRS-to-MODIS factor per country on Nov 2024 – Jul 2025. On five held-out months (Aug – Dec 2025) disagreement with MODIS fell in all four regions; in Ireland, for example, the error dropped from 3.8 to 1.2. The result is a 312-month calendar (Nov 2000 – Oct 2026) for Ireland, the UK, Spain and France, where any month can be compared with its historical median, with CSV/SVG export. Around this history we added a 24-hour fire-detection probability map built with NOAA GFS weather, wind-driven spread scenarios for Ireland, and a low-cost ESP32 temperature sensor that streams readings to the map. This matters because early-warning teams need to know whether a month is truly unusual, or only measured differently.

    Read moreShow less

    Team: Kyrylo Tonkovyd, Rodion, Amiliia Aziz kyzy, Matvii, weishun aw

    • 2026

    Flame Nebula

    Harmonization of MODIS and VIIRS Hot Spots

    Satellites have mapped wildfires since 2000, but the record is split between two witnesses: MODIS (Terra, Aqua), nearing retirement, and its successor VIIRS (Suomi NPP, NOAA-20, NOAA-21). VIIRS sees smaller fires, so the raw counts of the two sensors can't be compared, and long-term trends break where one hands over to the other. We developed Flame Nebula, a web app that harmonizes both records into one consistent 25-year fire history and makes it usable in seconds. A user picks any area (a region by name, or a shape drawn on the map) and gets a burning activity calendar, area statistics, year-to-year comparison, recurring fire patterns and downloadable data. This answers the challenge directly: the sensor of origin no longer matters, and patterns, trends and anomalies become the focus. It matters because emergency responders, firefighters, land managers and scientists need to know when and where fire comes back, and a broken record hides exactly that.

    Read moreShow less

    Team: Oleksandr Podynihlazov, Vladislav Onishchenko, Oleksandr Motoryhin, Oleksandr Kurilets, Daniel Kramnov, Dmytro Zubal

    • 2026

    FireBridge: One Fire Record from Two NASA Satellites

    by UCD Space Team

    Harmonization of MODIS and VIIRS Hot Spots

    FireBridge harmonises NASA's two wildfire-detecting satellite records, MODIS (since 2000, ~1 km pixels) and VIIRS (since 2012, 375 m pixels), into one consistent 25-year fire record. Because VIIRS sees far more small fires, simply combining the records creates a false jump in fire activity in 2012, and MODIS retires in 2027, so a consistent record is needed now. We processed 15.5 million NASA FIRMS detections: cleaning low-confidence and static heat sources, gridding them into half-degree daily cells, and learning a region-by-month correction from the years both sensors overlap. Tested on unseen years, this reduced the MODIS–VIIRS bias from −86% to −0.4%. On this record, an LSTM model forecasts fire activity and beats climatology and persistence baselines. Our web app serves fire managers, land managers and residents with a global fire overview, a historical fire calendar, forecasts with uncertainty bands, and CSV/PDF downloads. We also built a tabletop "mini satellite" — two infrared sensors sweeping a 3D-printed terrain — to demonstrate the problem physically and show tracking continuing after MODIS is switched off.

    Read moreShow less

    Team: SeungYeon Shin, Ignacio Salto Galdon, patrick ryan, Evan McMonagle, Li An Ma, Tao Luo