Looking at the GUARDIAN system’s “brain”: The NetSense SIDEINFO platform
Wildfire in Venta del Moro, 4th July 2022
In this article we show the main features of the monitoring, control and intelligence system that Medi XXI has designed for the GUARDIAN project. This is an extraordinary piece of engineering which includes several Industry 4.0 methods and tools and yet, very intuitive and user-friendly!
Internet of Things, Big Data, Machine Learning, etc. are on today’s agenda in many engineering projects of all sorts, and ours is not an exception. GUARDIAN takes benefit of several types of information technologies to provide a fully digitalized and interconnected service relying on hundreds of sensors deployed at the wildland-urban interface of Riba-roja del Túria and Paterna. Control and simulation algorithms are feed 24 hours a day to inform fire risk managers and first-responders from both municipalities with the current status and reliable forecasts in case of an emergency. Let me guide you through all the bits and pieces of this amazing system!
The NetSense at a glance
The GUARDIAN control and monitoring system is displayed through the NetSense platform by which several modules are integrated and visualized. Starting with a very intuitive home screen, the user can navigate through Riba-roja del Túria or Paterna municipalities to find information on weather, biomass and environmental parameters and wildfire danger. Moreover, NetSense is connected to a SCADA system to manage the hydraulic infrastructure (tank levels monitoring and control, irrigation for equipment maintenance, etc.) and to the specific module for activating the water canyons in case of fire emergency.
In Figure 1 the initial dashboard for Riba-roja del Túria is depicted. On-site general fire weather data together with the fire danger index are shown on the horizontal panel on top of the screen (look at these values of a chilly winter morning! quite different of what we would see in July…). These can be confronted with the information provided by Aemet, the Spanish agency of meteorology (blue vertical panel at the right hand side). The map at the centre powered by mapbox allows you to navigate through the area and display current sensor measurements (we will go through this soon..!) and finally, on your left, the vertical menu points at all the modules to control and manage the system.
Wheather, biomass and environmental parameters
The NetSense system enables remote data gathering of a dense network of sensors that has been deployed in the municipalities of Riba-roja del Túria and Paterna. There are three different types of monitoring stations equipped with several sets of devices to monitor the following variables:
Meteorological stations: temperature, relative humidity, atmospheric pressure, rainfall, wind speed and direction and solar radiation.
Environmental stations: temperature and relative humidity, NO2 and O3 concentration and particulate matter (PM10, PM2.5).
Biomass stations: temperature, relative humidity, soil moisture, soil temperature and foliar moisture.
Five meteorological stations are set in Riba-roja del Túria, distributed in key points (i.e. at the centre of the “Masia Traver” community, at the “Masia Traver” irrigation area, at the Camp de Túria II water treatment plant, at “Els Pous” water tank and at “Els Pous” irrigation area) while 5 more are set in Paterna (i.e. “Cañada Norte” and “Cañada Sur” residence areas and irrigation areas). The location was established with the requirement of evenly covering the overall study area and considering power supply needs (Figure 2).
As we all know, the role that weather plays in wildfire occurrence and behaviour is paramount, this is the reason why in GUARDIAN we look very closely to that!
There are three environmental stations providing data on air quality: one in “La Vallesa” forest (Paterna) and the other two in Riba-roja del Túria (one at “Camp de Túria II” and the other at “Els Pous”. As previously mentioned, they monitor the presence of particulate matter (inhalable particles with very fine diameter (PM10 particles have diameters < 10 micrometres and PM2.5 < than 2.5) that pollute the air coming from smokestacks, fires, unpaved roads, etc. Nitrogen oxide concentration in the atmosphere (NO2) is also obtained, which also gets in the air from the burning of fuel. Finally, ground-level ozone (which can be created from pollutants emitted by cars, power plants and other sources reacting in the presence of sunlight) is also sensed. GUARDIAN cares about air quality all-year round, even if it is not specifically wildfire-induced. To know more about air pollutants and their health effects there are no better sites as those from the European Environment Agency and the Environmental Protection Agency from the USA.
Biomass stations (four deployed in Paterna and two more in Riba-roja del Túria) provide basic information on the key indicators needed to monitor the water content in Riba-roja del Túria and Paterna forests. Indeed, data sensed by these stations feed directly the algorithm designed by Universitat Politècnica de València to activate preventive irrigation. FYI: the system is not running in automatic mode yet, but it triggers a virtual NetSense alarm when this need is detected.
One of the most valuable information that NetSense can show is the clear effect of the irrigation program on lowering the wildfire danger through the calculation of the Fire Weather Index (FWI) in Riba-roja del Túria and Paterna sites. Based on fire weather and fuel moisture data, FWI is a worldwide-used index that accounts for the effects of fuel water content and wind on wildfire behaviour and spread. Daily and seasonal FWI values can be consulted and retrieved from Copernicus Emergency Management Service at European level with 8 km resolution, which is good for an overall analysis of wide areas and regions. Nevertheless, the level of detail and precision that GUARDIAN achieves through its highly instrumented stations is astonishing.
In Figure 3, the evolution of the FWI calculated from June 22nd 2022 to July 12th 2022 in all monitored points in Riba-roja del Túria retrieved from the NetSense platform is shown. The green curve, with lowest FWI, corresponds to the area of “Els Pous” were the prescribed irrigation plan was implemented during that period. The effectiveness of GUARDIAN is undeniable!
Hydraulic infrastructure and activation of water canyons
NetSense is connected to the SCADA (Supervisory Control and Data Acquisition) system designed to control and manage the hydraulic infrastructure, which is run by HIDRAQUA, since the company works as local water utility. This module provides an overview of the general hydraulic diagram and detailed information of every piece of equipment (tanks, valves, pumping instrumentation, auxiliary energy supply systems, etc.) as shown in Figure 4 for BRAVO water tank set-up in “Cañada Norte” (Paterna).
This module is of course interconnected to the water canyon triggering system that allows the different types of irrigation patterns to be activated. As reported in past web articles, the water might be delivered in GUARDIAN WUI areas at different stages and for different purposes: preventive irrigation can be programmed to increase fuel moisture anytime, particularly when the area is experiencing rainfall deficit; pre-suppression wetting is used for fuel blanketing during the preparedness stage in case of fire in the area and, finally, direct attack might be started when the flames are so close that can be directly wet by the water delivered by the GUARDIAN canyons.
The way the system runs is very intuitive. Every site equipped with water canyons has an individual control screen (as the one for “Masia Traver” shown in Figure 5) by which irrigation can be remotely switched on and off.
The recommended operation of the system in case of a fire emergency is based on the definition of 3 areas of influence around every irrigation zone that are pre-set by default: area A where the dwellings to be protected are, area B being the region adjoining the houses, and area C which is further away in the municipality limits. These areas can be even worked out through fire behaviour simulations run with real conditions.
Indeed, NetSense has integrated a fire simulator module (Figure 6), based on the popular tools Behave and Flammap, developed by the USDA Forest Service. The resolution and accuracy of these tools running in our study site are as good as you can get, thanks to the fine job done by the GUARDIAN consortium while describing and modelling fuels with the LiDAR technology!
Now, if a fire is declared in area C of any irrigated zone, the operator of the NetSense platform can activate the canyons (see the green button in Figure 5). If he/she does, the irrigation system of this particular zone will be on up to consume the 25% of the total water tank capacity. If the fire gets into area B, the operator can activate again the irrigation system (see the orange button in Figure 5) and in this case, another 25% of capacity of the tank will be delivered. Finally, when/if the fire reaches area A, the operator can activate again the system (red button), which will deliver all the water remaining in the tank (the lasting 50%).
This operation will be done in close collaboration with the fire-fighters that will command the emergency and will have real-time information on the location and evolution of the fire perimeter.
The NetSense platform is currently being used by the municipalities of Riba-roja del Túria and Paterna to receive real-time data on fire weather parameters and air and fuel conditions. The system is operated by the local fire managers, by which they are promptly and precisely informed to take the best decisions concerning irrigation planning. Although NetSense was already operational during the summer of 2022, fortunately no emergencies occurred by then, so its performance could not be evaluated through a real fire scenario.
There is a still a way to go to have NetSense running in optimum mode. Of course, there are always algorithms that could be improved and different functionalities that could be implemented to automate processes and integrate all first-responders protocols, but the tool has been conceived with large flexibility to eventually account for that.
Furthermore, the potential of exploitation of the database associated to NetSense is just huge. We have to bear in mind that information is nowadays being sensed and stored at a frequency of 1 datum/minute. If we think of all the sensors deployed in Riba-roja del Túria and Paterna (164 sensors to be precise!) which are working in continuous mode, it gives us a massive amount of data. With proper Artificial Intelligence methods, these data could be used to make climate projections, mid-term FWI forecasts, and to fine-tune, for instance, the irrigation algorithm and the fire simulator model. I would like very much to encourage NetSense users to acknowledge this potential and take action! And rememeber…