Unity of satellite and weather data to battle fires
Posted on 20 Jul 2016
Meanwhile space dimension technologies highly evolve and artificial space satellites explore other planets for decades already, satellite technologies for observing the surface of our planet also become more complicated. With analysis of data received from satellites one can if not solve entirely then as minimum contribute to solving of global issues in many aspects of human life such as agriculture, logistics on ground surface and in open water, environmental pollution, etc.
And one of such global issues which has become extremely crucial during recent months is a fire. It happens in both hemispheres of the earth: in California, Arizona, Ontario, at Baikal Lake. Surely the main battle with fire takes place on earth surface, however such technologies as remote monitoring and other tools which provide timely data can and already do make performing of work tasks for firefighters easier. Generally speaking, unity of field maps and remote monitoring is a modern standard for fire guiding services. So how can remote monitoring help in battle with fires? The main questions that are of great importance in the beginning are whether there is a real fire or not, the amount of victimized territory, how much stuff have already burnt and how much can be burnt in the future. At this stage timely and proper operational data are crucial since these data are used in calculating how many human and technological resources are necessary to battle fire. So how can satellite data help then, and what can be calculated with its help? Nowadays there are several methods of detection of fires. All the satellites have infrared sensors, and open fires are well seen in infrared dimension as they beam as bright red points in this spectrum. Thus with the help of satellite monitoring any deviance is traced which likely can turn to be a fire. Unfortunately if a starting fire is small in size then it can’t be detected on images from satellites.
One more problematic aspect here is that one cannot definitely say that it is a fire for sure on an image, as a factory pipe or a roof highly heated under sun can also glow brightly in infrared spectrum. That is why the process of interpretation of these data is rather complicated and however there are such products as Fire Information for Resource Management System (FIRMS)by NASA, this question of interpretation of data is not completely solved yet.The second issue here is that a satellite is a little help in detection of unseen fire. It happens when peatbogs burn. If a latent fire is covered with leaves or peat burns under the surface of the ground, then such fires are hardly seen in infrared spectrum.
But in case when a fire affects big territories of hundreds of hectares, it is properly to speak about accurate evaluation of a situation with the help of satellite data. And in a situation when thousands of hectares are affected the remote monitoring provides the better perception of a case in comparison with ground research or a flight over a fire as there are risks of poor visibility, smoke and lack of fuel.Also satellite can help to detect the areas where smoke spreads. Smoke consists of aerosol particles and thus it can be easily differentiated from clouds, but not so easily from greenhouse gases coming with wind from neighboring fields.Summarizing it all, it becomes clear that at least at a current stage of technological progress the problems of timely fire detection and of getting accurate data about existing fires cannot be entirely solved using only satellite technologies. And here weather data come to help.
Combined analysis of satellite and weather data contributes to more complex and precise products. One can perform forecasting of behavior of points with starting fire: whether a fire is going to a halt or whether more territory can be affected. For instance, analysis of NDVI index provides knowledge that it is exactly wood burns. Further one can determine a degree of humidity and find out that this wood is dry, in a case. By analyzing available historical weather data it is possible to get to know that it has not rained recently, thus at the moment this forest is quite arid. Or vise versa, it has rained recently, air humidity is high and thus one can come to a conclusion that a fire will not last long.With forecasts for a specific location one can determine whether a storm with lightning is expected, as lightning on its own can cause fire and boost burning in a dry forest. Also forecasts show whether heavy rains and showers are possible during next 2 weeks, which can stop fire.
Therefore, there are several factors for analysis: the presence/absence of forest, the amount of fresh and dry leaves, humidity, forecast expectation of rains and storms. Considering these and other data a conclusion can be made whether the given territory is at risk or not for the next 2 weeks.
Later on, with knowledge of wind direction it is possible to predict spreading of fire and smoke, to determine presence of towns and settlements in affected areas, and to define degree and time frames of approaching danger and to warn inhabitants.
Unfortunately in cases when smoke covers large cities there is a little chance to battle fire with human-made tools. Authorities can try to stop fire spreading, can provide people with respirators, can involve air conditioners, can engage humanitarian aid. But only weather condition with heavy rains is in force to bring fire to a halt.Thus timely detection of fires, elimination of it at stages when it is possible to do or at least not using heavy machinery is important not only for firefighters who work on the ground, but also for inhabitants who live in potentially dangerous areas.