Search

Every year, thousands of forest hectares are burned worldwide, causing important consequences on the atmosphere, biodiversity and economy. Proper prediction of the fire evolution allows to manage the fire fighting equipment properly. Therefore, it is crucial to use reliable and fast simulations in order to predict the evolution of the fire. WRF-SFIRE is a wildland fire simulator, which couples a meteorological model called Weather Research and Forecasting Model (WRF) and a forest fire simulator, SFIRE reproducing the interaction between the propagation of the fire and the atmosphere evolution. The mesh resolution used to solve the atmosphere evolution has a deep impact in the prediction of small scale meteorological effects. At the same time, the ability of introducing these small scale meteorological events into the forest fire simulation implies enhancements in the quality of the data that drives the simulation. Therefore, better fire propagation predictions are obtained. However, this improvement can be affected by the instability of the problem to be solved. This paper states an accuracy study changing the mesh resolution when running WRF-SFIRE using as example a real case that took place in Catalonia (northeast of Spain) in 2005.

Many scientific works have focused on developing propagation models that predict forest fire behavior. These models require a precise knowledge of the environment where the fire is taking place. Geographical Information Systems allow us determining and building the different information layers that define the terrain and the fire. These data, along with meteorological information from weather services, enables the simulation based on real conditions. However, fire spread prediction models require a set of input parameters that, in some cases, are difficult to know or even estimate precisely. Therefore, a framework, based on a genetic algorithm calibration stage, was introduced to reduce the uncertainty in the input parameters and improve the accuracy of the predictions. This stage is implemented using a MPI master/worker scheme and an OpenMP parallel version of the fire spread simulator. Additionally, the whole system is run using suitable automatic worker-assignment and core-allocation policies to respect the existing time restrictions, inherent to this real-world problem. This paper details the process of obtaining the necessary input data as well as the parallel evolutionary framework that delivers the final prediction. A real case study is presented to illustrate the way this framework works.

In order to use environmental models effectively for management and decision-making, it is vital to establish an appropriate measure of confidence in their performance. There are different ways and different methodologies to establish and measure the confidence of the models. In this paper, we focus on the forest fire spread prediction. Simulators implementing forest fire spread models require diverse input parameters to deliver predictions about fire propagation. However, the data describing the actual scenario where the fire is taking place are usually subject to high levels of uncertainty. In order to minimize the impact of the input-data uncertainty a Two-Stage methodology was developed to calibrate the input parameters in (1) an adjustment stage so that the calibrated parameters are used, and (2) the prediction stage to improve the quality of the predictions. Is in the adjustment stage where the error formula plays a crucial role, because different formulas implies different adjustments and, in consequence, different wild fire spread predictions. In this paper, different error functions are compared to show the impact in terms of prediction quality in DDDAS for forest fire spread prediction. These formulas have been tested using a real forest fire that took place in Arkadia (Greece) in 2011.

Forest fire propagation prediction is a key point to fight against such hazards. Several models and simulators have been developed to predict forest fire propagation. These models require input parameters such as digital elevation map, vegetation map, and other parameters describing the vegetation and meteorological conditions. Coupling wind field model and forest fire propagation model improves accuracy prediction, but increases significantly prediction time. This fact is critical since propagation prediction must be provided in advance to allow the control centers to manage firefighters in the best possible way. This work analyses WindNinja execution time, describes a WindNinja parallelisation based on map partitioning, determines the limitations of such methodology for large maps and presents an improvement based on adapting map resolution to accomplish execution time limitations.

The most important aspect that affects the reliability of environmental simulations is the un- certainty on the parameter settings describing the environmental conditions, which may involve important biases between simulation and reality. To relieve such arbitrariness, a two-stage pre- diction method was developed, based on the adjustment of the input parameters according to the real observed evolution. This method enhances the quality of the predictions, but it is very demanding in terms of time and computational resources needed. In this work, we describe a methodology developed for response time assessment in the case of fire spread prediction, based on evolutionary computation. In addition, a parallelization of one of the most used fire spread simulators, FARSITE, was carried out to take advantage of multicore architectures. This al- lows us to design proper allocation policies that significantly reduce simulation time and reach successful predictions much faster. A multi-platform performance study is reported to analyze the benefits of the methodology.

Forests fires are a significant problem especially in countries of the Mediterranean basin. To fight against these disasters, the accurate prediction of forest fire propagation is a crucial issue. Propagation models try to describe the future evolution of the forest fire given an initial scenario and certain input parameters. However, the data describing the real fire scenario are usually subject to high levels of uncertainty. Moreover, there are input parameters that present spatial and temporal variation that make the prediction less accurate. Therefore, to overcome such uncertainty and improve accuracy it is necessary to couple complementary models such as the case of the wind field model. Such models use the meteorological forecasted wind to provide the wind direction and speed depending on the topography of the terrain. We use WindNinja as wind field simulator. This simulator takes a lot of time to deliver the predictions and it is a serious problem because fire propagation prediction must accomplish strict time constraints. To solve this problem, we propose map partitioning and solving independently for each one of the parts. However, the model has problems concerning boundary effects which is an additional source of uncertainty. Therefore, it is necessary to apply certain degree of overlapping among parts to reach a stable wind field without inconsistencies and a minimum uncertainty.

The accurate prediction of forest fire propagation is a crucial issue to minimize its effects. Several models have been developed to determine the forest fire propagation. Simulators implementing such models require diverse input parameters to deliver predictions about fire propagation. However, the data describing the actual scenario where the fire is taking place are usually subject to high levels of uncertainty. The input-data uncertainty represents a serious drawback for the correctness of the prediction. So, a two-stage methodol- ogy was developed to calibrate the input parameters in an adjustment stage so that the calibrated parameters are used in the prediction stage to improve the quality of the predictions. This way, we relieve the effects of such uncertainty. In this work, we take advantage of this two stage methodology applying Genetic Algorithms as the calibration technique. However, the use of Genetic Algorithms require the execution of many simulations. This fact, added to the eventual long executions of the underlying simulator (due to its inherent complexity), implies to deal with another serious problem: the time needed to deliver the predictions. To be useful, the prediction must be provided much faster than real time. So, it is necessary to exploit all available computing resources. In this work, we present a two-stage forest fire spread prediction hybrid MPI-OpenMP application based on the Master- Worker paradigm and the parallelization of the FARSITE simulator in order to minimize the response time. The results as regards the enhancement in the quality of the predictions are reported, as well as the results regarding the time saving obtained by this hybrid application.

A quick response becomes crucial in natural hazard management. When an emergency occurs, hazard evolution simulators are a very helpful tool for the teams in charge of making decisions. To perform the simulations, they rely on data which usually constitutes a big set of parameters, which have been previously recorded from observations, usually coming from remote sensors, images, etc. However, this data is frequently subject to a high degree of uncertainty. This data uncertainty also produces uncertainty in simulators’ results. To overcome this drawback, in previous works we developed a two-stage prediction method, which has been demonstrated to improve noticeably the quality of the predictions. The time incurred in performing this strategy, however, may vary significantly depending on different factors. As it is well known, the execution time of a particular simulator depends on the specific setting of the input parameters. Moreover, decision control centers in charge of making decisions to fight against the ongoing disaster require a certain degree of quality in the final prediction. Depending on how demanding are the quality and time constraints, the two-stage strategy may need the support of Urgent Computing solutions, in order to meet the requirements. In this work, we focus on forest fires spread prediction, as a real application case of study, and we expose our methodology to characterize both the time needed and the expected quality of the two-stage prediction method, so that we are able to determine the amount of computational resources needed to respond effciently to an eventual emergency. This way, we emphasize the need of Urgent Computing mechanisms to be able to put into practice this method at real time.

Natural hazards are significant problems that every year cause important loses around the world. A good prediction of the behavior of the hazards is a crucial issue to fight against them and to minimize the damages. The models that represent these phenomena need several input parameters and in many cases, such parameters are diffcult to know or even to estimate in a real scenario. So, a methodology based on the DDDAS paradigm was developed to calibrate the input parameters according to real observations of the behavior and evolution of the hazard. Such calibrated parameters are then used to provide an improved prediction for the next time interval. This methodology was tested on Forest Fire Propagation Prediction with significant results. The developed methodology takes the fire behavior and propagation during a time interval and then searches for the values of the input parameters that best reproduce the propagation of the fire during that interval. Several Artificial Intelligence (AI) methods were applied to carry out this search as fast as possible. The values of the parameters that best reproduce the behavior of the fire were then used as input parameters to predict the propagation during the next time interval. These parameters were considered constant during both time intervals and a single value for each parameter was used for the calibrating process and for the prediction stage. This methodology fits on the DDDAS paradigm since the prediction is dynamically driven by the system evolution. However, there are several parameters that are not constant through time, but they may vary dynamically. In the case of forest fires, a typical example is the wind. In some cases, when the time interval is short an average value for the wind can be a feasible value, but when the time interval is longer, in most cases, a single value cannot represent the variability of the wind. We can estimate wind behavior applying some complementary model. In this work, we are going a step further considering the dynamic behavior of such parameters. We propose an extension of the existing prediction scheme that takes into account the dynamically changing parameters by coupling a weather prediction system on a DDDAS Forest Fire Propagation Prediction system.

When an emergency occurs, hazard evolution simulators are a very helpful tool for the teams in charge of making decisions. These simulators need certain input data, which defines the characteristics of the environment where the emergency is taking place. This kind of data usually constitutes a big set of parameters, which have been previously recorded from observations, usually coming from remote sensors, pictures, etc. However, this data is frequently subject to a high degree of uncertainty, as well as the results produced by the corresponding simulators. Hence, it is also necessary to pay attention to the simulations’ quality and reliability. In this work we expose the way we deal with such uncertainty. Our research group has previously developed a two-stage prediction methodology that introduces an adjustment stage in order to deal with the uncertainty on the simulator input parameters. This method significantly improves predictions’ quality, however, in order to be useful, a good characterization of the adjustment techniques has to be carried out so that we are able to choose the best configuration of them, given certain restrictions regarding resources availability and time deadlines. In this work, we focus on forest fires spread prediction as a real study case, for which Genetic Algorithms (GA) have been demonstrated to be a suitable adjustment strategy. We describe the methodology used to characterize the GA and we also validate it when assessing in advance the quality of the fire spread prediction.

The need for input parameter optimisation in environmental modelling is long known. Real-time constraints of disaster propagation predictions require fast and efficient calibration methods to deliver reliable predictions in time to avoid tragedy. Lately, evolutionary optimisation methods have become popular to solve the input parameter problem of environmental models. Applying a knowledge-guided Genetic Algorithm (GA) we demonstrate how to speed up parameter optimsation and consequently the propagation prediction of environmental disasters. Knowledge, obtained from historical and synthetical disasters, is stored in a knowledge base and provided to the GA in terms of a knowledge chromosome. Despite of increased loads of knowledge, its retrieval times can be kept near-constant. During GA mutation, ranges of selected parameters are limited forcing the GA to explore promising solution areas. Experiments in forest fire spread prediction show how time-consuming fitness evaluations of the GA could be lowered remarkably to cope with real-time capabilities maintaining the error magnitude.

Forest fires cause important losses around the world every year. A good prediction of fire propagation is a crucial point to minimize the devastating effects of these hazards. Several models that represent this phenomenon and provide a prediction of its spread have been developed. These models need input parameters which are usually difficult to know or even estimate. A two- stage prediction methodology was proposed to improve the quality of these parameters. In this methodology, such parameters are calibrated according to real observations and then, used in the prediction step. However, there are several parameters, which are not uniform along the map, but vary according to the topography of the terrain. Besides, these parameters are not constant along time but they are strongly dynamic. In such cases, it is necessary to introduce complementary models that overcome both restrictions. In the former case, the need of a spatial distribution model of a given variable is needed to be able to provide a spatial distribution for a given variable along the whole terrain by starting from the measured values of that parameter in certain points of the terrain. In the case of time variability, a complementary model such as weather forecasting model, could enable the capability of dealing with dynamic behavior of these parameters along time. In this paper, we describe an enhanced two-stage prediction scheme, where both type of complementary models a wind field model and a weather prediction model are coupled to the prediction scheme by enabling the system to dynamically adapts to complex terrains and dynamic conditions.

This work faces the problem of quality and prediction time assessment in a Dynamic Data Driven Application System (DDDAS) for predicting natural hazard evolution. In particular, we used forest fire spread prediction as acase study to show the applicability of the methodology. The improvement on the prediction quality when using a two-stage DDDAS prediction framework has been widely proved. The two-stages DDDAS has a first phase where an adjustment of the input data is performed in order to be applied in the second phase, the prediction. This paper is focused on defining a new methodology for prediction time assessment under this kind of prediction environments by evaluating, in advance, how a certain combination of simulator, computational resources, adjustment strategy, and frequency of data acquisition will perform, in terms of prediction time. Since the time incurred in the hazard simulation is a crucial part of the whole prediction time, we have defined a methodology to classify the simulator's execution time using Artificial Intelligence techniques allowing us to determine upper bounds for the DDDAS prediction time depending on the particular input parameter setting. This methodology can be extrapolated to any DDDAS for predicting natural hazards evolution, which uses the two-stage prediction scheme as a working framework.

Data assimilation in weather forecasting is a well-known technique used to obtain an improved estimation of the current state of the atmosphere (analysis). The Meteorological Service of Catalunya (SMC) is seeking for a real time high resolution analysis of surface parameters over Catalonia (north-east of Spain), in order to know the current weather conditions at any point of that region. For this purpose, a comparative study among several data assimilation experiments based on LAPS (Local Analysis and Prediction System) and STMAS (Space-Time Multiscale Analysis System) and multi-regression technique designed at SMC, has been performed to determine which one delivers best results. The comparison has been done using as true state independent observational data provided by the Spanish Meteorological State Agency (Agencia Estatal de METeorologia, AEMET). The results show that the multi-regression technique provides more accurate analyses of temperature and relative humidity than the LAPS/STMAS experiments, mainly due to the fact that multi-regression methodology only uses observations and consequently the model biases are avoided.