Aerial Fire Tech‎ > ‎


Infrared light and colors used for mapping fires

Color is the only light human eyes see but there are many others. Thermal infrared (IR) is heat emitted as light. It shines through smoke better than sunlight through clear water, but like sun through water, too much smoke will eventually cloud out the thermal light. Fire puts out a lot of lights, some we see, some we feel as heat. Thermal comes as either middle wave (MWIR) or long wave (LWIR).   Color, Near InfrRed (NIR, used for night security cameras) and Short Wave (SWIR) are not thermal bands, but cameras with lens appropriate to the band, can focus any of them into a picture.

From iPhones to fancy sensors, 8-bit pictures provide 256 shades and is the norm. 16-bit seems only twice 8, not. It's 256 times more. Our eyes see only ~25 light to dark shades of any one color, and 8-bit's 256 shades seem like a lot, not. Every 65,535 shade of 16-bit imagery is a temperature increment on our thermal radiometric (temperature calibrated) cameras. That's .04 degrees. At 25,000 ft, tiny fire starts might only discolor a pixel by .04 degree/shade.

LWIR and MWIR both map fires well. I prefer LWIR due to low cost, low power, simpler, smaller, wide dynamic range and rugged. LWIR or MWIR, 14 or 16-bit is the best for fire mapping and allows radiometric resolution (temperature readout).  Medium Wave IR (MWIR) is good too. Some like it for clarity due to a narrower dynamic range. High clarity can also be produced by image processing LWIR.  MWIR requires a cooler and has more zoomable 
telephoto lens options that magnify distant objects, a feature highly valued for surveillance but not needed for fire mapping. Fire lens are wide-angle and fixed (changing zoom disrupts stitching mosaics). 

Long Wave Infrared (LWIR) 7-14 micron. The best for firefighting. It shows mountains, fire and vegetation through thick smoke and dark. 16-bit has 65535 shades of grey which equate to temperatures ranging from 0F to over 1000F in .04 increments, good dynamic range. Older microbolometers have more noise but new ones are much less.  Beypond LWIR is Microwaves used for ovens, radar and radio communications. LWIR sees through smoke and haze the best. It will see through thin clouds but not thick ones. 

Middle Wave Infrared (MWIR) 3-7 micron. Known for being crisper than LWIR in 8-bit images, but 14-bit contrast is about the same with less dynamic range. Much more expensive, larger and troublesome (requires a sensor cooler refrigeration). Data is usually less noisy but new LWIR is matching it.  LWIR and MWIR both see through smoke, haze and thin clouds but not thick ones.

14/16-bit thermal (MWIR or LWIR) allows the same photo or video to show detail in both cold ground vegetation adjacent to detail of high heat fire. In 8-bit will show one or the other. 14-bit can be re-processed (like PhotoShop) to show different details not visible to the human eye in real-time, like fire bloom. Bloom is the bright haze surrounding intense fire heat, but this can be eliminated with image processing of 14-bit imagery. 14/16-bit thermal see heat about as well during the day as 8-bit sees heat at night. 8-bit is much less good during the day.

Short Wave Infrared (SWIR) 1.1-3 micron/1000-3000 Nanometers. Can be slightly useful but limited and mostly redundant. Used occasionally for visualizing fire and ground and vegetation along with fire during the day. At night the reflective sunlight is absent, and thermal is not as good as LWIR and MWIR. Not best for mapping.  Very good for oil spill mapping.

Near-Infrared (NIR).  .8-1.1 Micron/800-100 Nanometers. Not relevant to firefighting. Used for plant analysis during the day. Does not work at night or through smoke, but does see through haze. Can be helpful sometimes when mapping the black during the day. Some people can see the lower range with their eyes, and modern cameras all see NIR, and a cut filter is used to eliminate

Color, Highest resolution and least expensive of all the cameras. Red-Green-Blue (RGB), Yellow Red Blue primary colors same as a rainbow, Cyan-Magenta-Yellow-Black (CMYK). Additive and subtractive colors. RGB vs YRB vs cyan...
red edge not relevant to fire fighting provides analytics for vegetation and agriculture

Ultraviolet (UV) Corona. Sometimes can be seen as a band in the rainbow. This light is near x rays and creates sunburns. Used for seeing electric failures and oil spills

Hybrid is a function by gimbal makers that blends lower resolution IR with high resolution color to create a more ascetically pleasing image

NDVI: Normalized Difference Vegetation Indexes are used for analyzing plant health, especially for agriculture. There are many Vegetation Indexes depending on what is being analyzed.  Not relevant to fire, but could help map black during the day

CIR: Color with Near-Infrared blended to better show vegetation. Not relevant to fire unless mapping the black during the day.


Cameras are typically either in gimbal, also known as a ball or FLIR ball. Or are fixed mounted as multiple same cameras for wider swath, or different camera for multispectral. Hyperspectral is many colors/IR.


LiDAR radiates energy (shines like a flashlight) and reads the reflection for physical mapping.


Interesting point: Humans have infrared sensors like snakes. All of our skin senses thermal IR, but our faces are the most sensitive.  Between our eyes is higher sensitivity, like a snake but snakes are even more sensitive and partially focus the thermal heat.