Don’t really know where to put this thought. Now that the State is covered in smoke, what kind of detection devices can be used if another lightning storm comes through to see the incipient fires before they create a bubble on top of the smoke layer?
IR sensors can usually see through smoke. It is moisture in clouds that
attenuates heat signals in that part of the spectrum. So worry about cloud cover more than smoke for the NIROPS platform and satellite based sensors such as MODIS. Mountaintop cams will still struggle with smoke in the visible spectrum.
7 Likes
How big does the heat signal have to be to detect it? Time on a new start is critical. Calfire was set up for a twenty minute response for air support and stations are dispursed statewide to achieve a prompt initial attack on a new start.
Like most things, it depends. IIRC the NIROPS airborne ships flying ~10,000’ AGL should be able to detect a hotspot less than a meter in size, depending on actual source temperature, angle of input, atmospheric conditions, etc. Of course this is not designed to be used for initial attack response due to time needed to process the data. You can look up specs for helicopter mounted FLIR (Forward Looking Infrared) units which would be more real-time utility for IA response. Not sure what the MODIS specs are but there is a wealth of online explanations for that.
2 Likes
CalFire and USFS have a partnership with the National Guard for a high temporal and somewhat high spatial resolution sensor for fire detection.
We cant use modis as its not geostationary. Terra and Aqua pass over once a day with a wide swath of 2600km and has an exact revisit cycle of 16 days
Do you have a link to general info on the National Guard agreement? It has been a while since I have been able to keep up on all this IR stuff.
way above my pay grade I am not aware of what the agreement is only that it exists and to me just such a great use of of our tax dollar for a HADR effort.
1 Like
NIROPS (the Forest Service’s National Infrared Operation) has a policy of not conducting searches for lightning starts except as “spots” within boxes designated by a unit or IMTs as part of a “going-fire.” NIROPS is such a limited resource that they can’t do needle-in-a-haystack searches. When you understand the Phoenix line scanner it becomes clear why. I try to explain that below. Then I explain a little about satellite alternatives, which turn out to be much more effective for the detection mission.
The specs that NIROPS publishes are based on flying at 10,000 AGL. But, in fact, they more often fly at 15,000 because that allows the sensor to look over a wider area (at the cost of lowered ability to see small fires). Wider areas surveilled allow them to complete a request “quickly” and move to fill the order from the next fire. With their King Air and their Citation they have occasionally managed to fill 40 orders per night. To bag that many fires they can’t be very far apart or far from a source of Jet A.
Their scanners (called Phoenix) can be sampled about 1600 times in its cross-track sweep (think of this as your old analog TV sweeping the electron gun across the screen). At 15000 feet over flat ground, Phoenix samples a square directly under the aircraft that is 5.7 meters (18.7 feet) on a side. The ground sample distance boxes get bigger the farther off to the side of the aircraft you get. The scanner is sampled across a 120 degree arc under the plane (60 degrees either side of straight down). This gives a cross-track swath of about 14.2 kilometers (8.8 miles) wide. If the plane flies at 200 knots the scanner sweeps across 100,000 acres in an hour. And that is without turns etc. (The scanner can’t quite spin fast enough for 200 kts but for round number it is close). 100K acres is not very much area when you think about a thunderstorm going up the Coast Range.
The Phoenix sensor can detect an object above about 550 degrees C (1000 F, approximately the lowest flame temperature) that is in the range of 1/50th down to maybe 1/100th of its ground sample size. At 15,000 feet that is, at best, a 22 inch by 22 inch flaming object sitting on flat ground out in the open. With topography and canopy cover the question of what the Phoenix sensor can detect is a huge can of worms.
A few days ago, under the topic “MODIS not Updating” in the California Q&D section of this forum I wrote a post that included what the current GOES satellites and the MODIS satellites can detect. Because the sensor physics and the signal to noise ratios are in sort of the same ballpark, you can also expect the satellites to detect a flaming object that occupies 1/100th of the detectors ground sample size. For the 1 Kilometer x 1 Kilometer pixels of MODIS and the GOES Meso-Views that is about 2.5 acres.
In that other post I said that I had seen detectable signals from the GOES-West Meso view (nominally 1 km GSD) at the time when the first arriving units to a brush fire in So Cal reported “2 to 3 acres in moderate brush running up the hill”. To a small extent, such a report begins to square up the calculations and reality.
We have found both in theory and practice that — using data- and image-processing techniques which effectively increase the signal to noise ratio — we should be able to reliably detect fires that are in the 1 to 2 acre range under favorable conditions from the GOES Meso data.
It would help if you could detect a fire before is an acre running up a hill.
Carl Pennypacker and others from UC Berkeley published a paper in 2013 showing that a satellite in geo-stationary orbit with a 1/2 meter mirror (telescope) can (without image processing) detect a flaming area that is 10 feet by 10 feet. (UC holds the patent on this concept and practice. Fireball has exclusively licensed the patent.)
The Space Science Lab at Berkeley has a design for this instrument as a “hosted payload.” That means the payload would hitch a ride on a multi-use satellite. Considering that one such payload would surveill the western hemisphere for 20 years or more the cost of the payload, the ground components of the system, and the launch work out to about $4 Million per year of life. If this detector guided rapid suppression responses such a satellite would be a bargain.
6 Likes
Understand and concur. I think I have occasionally seen that product and just wondered. I imagine public info would be hard to come by.
There are a variety of assets within the “national technical means” which collectively are known as OPIR Overhead Persistent Infrared. (Google OPIR) Northrop Grumman recently got a $2.3B contract for the newest OPIR platform. The primary missions are related to missiles.
Not all of the assets are in orbit.
Most of the components themselves are publicly disclosed.
It is the capabilities and modes of operation that help keep us safe; and thus are classified.
4 Likes