A forest fire situation awareness system and method based on optical active and passive remote sensing data

Abstract

The application provides a forest fire situation awareness system and method based on optical active and passive remote sensing data, and relates to the technical field of natural disaster management.The system comprises an optical active remote sensing subsystem for acquiring real-time atmospheric observation data in the air above a target area, an optical passive remote sensing subsystem for acquiring real-time ground observation data of the target area, and a ground processing and modeling unit comprising a processor and a memory, wherein the processor is configured to perform the following steps: constructing an augmented state vector in the memory, running an ensemble of coupled models based on the augmented state vector to generate predicted values, calculating the residual between real-time observation data and the predicted values, and updating the vector, and running the coupled models based on the updated vector to generate final products. The multimodal collaborative data of optical active remote sensing and optical passive remote sensing technology is used to simultaneously estimate and adaptively optimize the internal key physical parameters and their change trends of a coupled physical model, thereby improving the prediction performance of the model.

Description

Technical Field

[0001] This application relates to the technical field of natural disaster management, and in particular to a forest fire situation awareness system and method based on optical active and passive remote sensing data. Background Technology

[0002] Forest fire situational awareness is central to forest fire management, requiring commanders to have an accurate grasp of the current state (perception), development mechanisms (understanding), and future trends (prediction) of the fire. However, existing technologies face two fundamental technical challenges in achieving high-fidelity situational awareness:

[0003] 1. Challenge 1: The fatal flaw in static combustible material databases

[0004] The predictive power of operational fire behavior models (such as WRF-Fire) largely depends on the accuracy of surface combustible data. Currently widely used combustible databases (such as LANDFIRE) have update cycles of several years, resulting in a serious disconnect between the information in the database (such as combustible type, load, and moisture content models) and the dynamic reality of forest ecosystems (such as drought and pests).

[0005] 2. Challenge Two: Loss of Control in the Fire-Gas Coupling Feedback Loop

[0006] The second major source of uncertainty in wildfire forecasting is the two-way coupling feedback between fire and the atmosphere. Large wildfires release enormous amounts of heat, generating strong updrafts and "fire-generated winds." While existing fire behavior models (such as WRF-Fire) theoretically couple atmospheric modules, the initial fields and boundary conditions of these modules are typically derived from standard mesoscale weather forecasts, and their spatiotemporal resolution is far from sufficient to resolve the dramatic local atmospheric variations at fire sites. Furthermore, the planetary boundary layer parameterization schemes built into WRF (such as YSU and MYNN) are designed for normal weather conditions, and they often underestimate the intense turbulent mixing driven by fire.

[0007] To address the aforementioned challenges, those skilled in the art have attempted to employ data assimilation techniques to incorporate real-time observation data into the model. For example:

[0008] (1) Status correction: The real-time fire line position (FRP or fire line perimeter) observed by the satellite is directly "corrected" or "covered" by the fire line status predicted by the model through data assimilation (such as the Level-Set function LFN in WRF-Fire).

[0009] (2) Atmospheric constraints: The atmospheric state variables of the model are "corrected" by assimilating the parameters (such as U, V, W) observed by active / passive technologies.

[0010] Existing solutions are limited to "state correction" to "static parameter estimation". This "state correction" approach has a fundamental flaw: it only corrects the "symptoms" of the model prediction (such as fire line location LFN error) without addressing the "cause" of the error, namely the systematic deviation of the model's internal physical parameters (such as combustible material moisture content model parameters or PBL turbulence parameters).

[0011] More advanced techniques in this field address this problem through "parameter estimation," which uses observational residuals to correct for the parameter values ​​themselves. However, while this "static parameter estimation" is superior to "state correction," it still suffers from a fundamental flaw: it assumes that the optimized parameters remain constant between two assimilation processes. This assumption is flawed in scenarios with strong fire-gas coupling. Fire is a highly dynamic process; the fire itself actively and continuously alters the parameters of its surrounding environment. For example, the thermal radiation from the fire continuously dries out the combustible material in front, causing a continuous decrease in the fuel moisture content (FMC) parameter; the enormous heat flux released by the fire continuously intensifies local turbulence, causing a continuous increase in the turbulent mixing length parameter of the fire-prone black-bone flow (PBL). Existing "static parameter estimation" cannot capture this "parameter change trend" driven by the fire itself, thus its "predictive skill" remains limited. Summary of the Invention

[0012] Therefore, it is necessary to provide a forest fire situational awareness system and method based on optical active and passive remote sensing data to address the above-mentioned technical problems. This system utilizes multimodal collaborative data from optical active remote sensing (especially atmospheric sounding lidar) and optical passive remote sensing (especially satellite imaging) technologies to synchronously estimate and adaptively optimize the internal key physical parameters and their changing trends of a coupled atmospheric-fire behavior physical model (such as WRF-Fire), thereby improving the model's predictive performance and generating high-fidelity situational awareness products for combat decision-making.

[0013] Firstly, this application provides a forest fire situation awareness system based on optical active and passive remote sensing data, comprising:

[0014] The optical active remote sensing subsystem includes at least one Doppler wind lidar for acquiring real-time atmospheric observation data over the target area; the real-time atmospheric observation data includes turbulent kinetic energy profiles or vertical velocity variance.

[0015] The optical passive remote sensing subsystem includes a satellite remote sensing platform for acquiring real-time surface observation data of the target area; the real-time surface observation data includes the calculation of real-time fire line positions using thermal infrared and shortwave infrared bands;

[0016] The ground processing and modeling unit includes a processor and memory, and the processor is configured to perform the following steps:

[0017] An augmented state vector is constructed in memory, and the ensemble of the coupled model is run based on the augmented state vector to generate model predictions. The augmented state vector includes the dynamic state variables of the coupled model, adjustable physical parameters of at least one atmospheric physics module, adjustable physical parameters of at least one surface physics module, and a parameter variation trend term for at least one of the adjustable physical parameters. The coupled model includes at least one atmospheric physics module and one surface physics module, including a WRF-Fire model containing an atmospheric module and a fire behavior module. The adjustable parameters of the atmospheric physics module include at least the turbulent mixing length or TKE closure constant in the planetary boundary layer parameterization scheme. The parameter variation trend term corresponding to the adjustable parameters of the atmospheric physics module includes the time rate of change of the turbulent mixing length. The adjustable parameters of the surface physics module include at least the equilibrium water content deviation term of the combustible water content time-delay model. The parameter variation trend term corresponding to the adjustable parameters of the surface physics module includes the time rate of change of the equilibrium water content deviation term.

[0018] Calculate the residuals between real-time atmospheric observation data, real-time surface observation data and predicted values, and update the augmented state vector based on the residuals using a data assimilation algorithm;

[0019] Forest fire situation awareness products are generated by running a coupled model based on the updated augmented state vector.

[0020] Secondly, this application also provides a forest fire situation awareness method based on optical active and passive remote sensing data, including:

[0021] An augmented state vector is constructed in memory, and a coupled model is run based on the augmented state vector to generate model predictions. The augmented state vector includes dynamic state variables of the coupled model, adjustable physical parameters of at least one atmospheric physics module, adjustable physical parameters of at least one surface physics module, and a parameter variation trend term for at least one of the adjustable physical parameters. The coupled model includes at least one atmospheric physics module and one surface physics module, including a WRF-Fire model comprising an atmospheric module and a fire behavior module. The adjustable parameters of the atmospheric physics module include at least the turbulent mixing length or TKE closure constant in the planetary boundary layer parameterization scheme. The parameter variation trend term corresponding to the adjustable parameters of the atmospheric physics module includes the time rate of change of the turbulent mixing length. The adjustable parameters of the surface physics module include at least the equilibrium water content deviation term of the combustible water content time-delay model. The parameter variation trend term corresponding to the adjustable parameters of the surface physics module includes the time rate of change of the equilibrium water content deviation term.

[0022] Real-time atmospheric observation data and real-time surface observation data are acquired. The residuals between the real-time atmospheric observation data, real-time surface observation data and predicted values ​​are calculated. The augmented state vector is updated based on the residuals using a data assimilation algorithm. The updated augmented state vector includes the updated adjustable physical parameters and the corresponding parameter change trend terms. The real-time atmospheric observation data includes turbulent kinetic energy profiles or vertical velocity variances. The real-time surface observation data includes the real-time fire line location calculated using thermal infrared and shortwave infrared bands.

[0023] Forest fire situation awareness products are generated by running a coupled model based on the updated augmented state vector.

[0024] In one embodiment, generating a forest fire situational awareness product by running a coupled model based on the updated augmented state vector includes:

[0025] Extract the updated adjustable physical parameters and corresponding parameter change trend terms from the augmented state vector after the update;

[0026] A dynamic parameter prediction model is constructed based on the updated adjustable physical parameters and the corresponding parameter change trend terms.

[0027] The dynamic parameter model is injected into the coupled model to generate an adaptively optimized coupled model, and the output of the adaptively optimized coupled model is used to generate a forest fire situation awareness product.

[0028] This application employs the above-mentioned forest fire situation awareness system and method based on optical active and passive remote sensing data, which has the following beneficial effects:

[0029] 1. The solution proposed in this application achieves a leap from "static estimation" to "dynamic trend inversion". It utilizes multimodal collaborative data from optical active remote sensing and optical passive remote sensing technologies, and introduces a high-order augmented state vector containing parameter trend terms (such as time derivatives) to achieve synchronous estimation and adaptive optimization of the internal key physical parameters and their changing trends of a coupled atmosphere-fire behavior physical model. This overcomes the limitation of existing parameter estimation techniques that assume parameters to be constant within the assimilation period, improves the predictive performance of the model, and generates high-fidelity situational awareness products for combat decision-making.

[0030] 2. Because the prediction model in this application uses dynamically changing parameters. Instead of static parameters, it can more accurately predict the acceleration or sudden change of fire, which is crucial for early warning of extreme fire behaviors such as "fire explosion" and greatly improves the prediction skills for "mutation-type" fire behaviors. Its beneficial effects far exceed those of static parameter estimation. Attached Figure Description

[0031] Figure 1This is a schematic diagram of a forest fire situation awareness system architecture based on optical active and passive remote sensing data in one embodiment;

[0032] Figure 2 This is a data flow diagram illustrating the core innovation of this application, the "dual augmented state vector parameter estimation" mechanism, in one embodiment.

[0033] Figure 3 This is a flowchart of a forest fire situation awareness method based on optical active and passive remote sensing data in one embodiment. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0035] The core innovation of this application lies in the first-ever proposed data assimilation method based on dual augmented state vectors, which is used to simultaneously estimate atmospheric physical parameters and surface combustible physical parameters in a coupled atmosphere-fire behavior model (such as WRF-Fire). The aim is to overcome the fundamental deficiency of existing technologies that can only perform "state correction" but cannot correct the physical parameters inside the model.

[0036] Firstly, referring to Figures 1 to 3 This application provides a forest fire situation awareness system based on optical active and passive remote sensing data, comprising: an optical active remote sensing subsystem, an optical passive remote sensing subsystem, and a ground processing and modeling unit. The optical active remote sensing subsystem is used to acquire real-time atmospheric observation data over the target area. O lidar The optical passive remote sensing subsystem is used to acquire real-time surface observation data of the target area. O sat The ground processing and modeling unit includes a processor and memory.

[0037] In this application, the processor is configured to perform the following steps:

[0038] S100, construct the augmented state vector in memory. Z aug Based on the augmented state vector, the ensemble of the coupled model is run to generate model predictions.

[0039] In this application, the augmented state vector Z aug Dynamic state variables of a coupled model Z state At least one adjustable physical parameter of atmospheric physics module Z param_atmAt least one adjustable physical parameter of the surface physical module Z param_fuel and at least one adjustable physical parameter's parameter change trend term Trend param The coupled model should contain at least one atmospheric physics module and one surface physics module.

[0040] For example, the standard dynamic state variables of a coupled model Z state It can provide wind field (U, V, W), temperature T, and fire line position function LFN, etc.; adjustable physical parameters of the atmospheric physics module. Z param_atm The turbulent mixing length in the planetary boundary layer (PBL) parameterization scheme can be used. L mix Adjustable physical parameters of the surface physics module Z param_fuel This includes adjustable parameters for surface combustible models, such as the equilibrium moisture content deviation in the FMC time-delay model. E fmc Trend of key parameters Trend param It can include at least L mix or E fmc The rate of change over time.

[0041] The prediction model in this application uses dynamically changing parameters. (), rather than static parameters; where, It is a state quantity that changes over time. Therefore, it can more accurately predict the acceleration or sudden change of fire, which is crucial for early warning of extreme fire behaviors such as "flame explosion" and greatly improves the prediction skills for "mutation-type" fire behaviors. Its beneficial effects far exceed those of static parameter estimation.

[0042] S200, calculates real-time atmospheric observation data O lidar and real-time surface observation data O sat The residuals between the predicted and actual values ​​are used to update the augmented state vector based on the residuals using a data assimilation algorithm.

[0043] The ensemble of the running coupled model (containing N members, each of which is perturbed) calculates the difference between the observed and predicted data sets, i.e., the residuals, at the assimilation time. For example, the atmospheric residual is the difference between the TKE profile observed by active remote sensing (lidar) and the TKE profile predicted by the model ensemble; the surface residual is the difference between the fireline location observed by passive remote sensing (satellite) and the fireline location predicted by the model ensemble.

[0044] A data assimilation algorithm, such as the Ensemble Kalman Filter (EnKF), is employed to update the entire augmented state vector based on the aforementioned combined residuals. During this process, the updated parameters include the adjustable physical parameters of the atmospheric physics module. Z param_atm The current value and adjustable physical parameters of the surface physics module Z param_fuel The current value; and, during the update process, the parameter change trend term of at least one adjustable physical parameter is inverted.

[0045] S300 generates forest fire situational awareness products based on the updated augmented state vector running coupled model.

[0046] In one embodiment, generating a forest fire situation awareness product by running a coupled model based on the updated augmented state vector includes: extracting the updated adjustable physical parameters and corresponding parameter change trend terms from the updated augmented state vector; constructing a dynamic parameter prediction model based on the updated adjustable physical parameters and corresponding parameter change trend terms; injecting the dynamic parameter prediction model into the coupled model to generate an adaptively optimized coupled model; and generating a forest fire situation awareness product based on the output of the adaptively optimized coupled model.

[0047] The system extracts the updated parameter values ​​and their corresponding trend terms to construct a dynamic parameter prediction model. This dynamic parameter prediction model (rather than static values) is injected into the physical scheme of the coupled model to adaptively optimize the coupled model for prediction in the next time step. The prediction results of the optimized coupled model then serve as the basis for generating forest fire situational awareness products.

[0048] In one embodiment, the optical active remote sensing subsystem includes at least one Doppler wind lidar. The Doppler frequency shift data acquired by the Doppler wind lidar is used to process atmospheric scattering signals to obtain turbulent kinetic energy (TKE) profiles or vertical velocity variance σ. w 2 As atmospheric observation data.

[0049] In one embodiment, the optical passive remote sensing subsystem includes a satellite remote sensing platform for acquiring surface observation data in real time. O sat This includes calculating the real-time fire line location using thermal infrared and shortwave infrared bands. LFN obs .

[0050] In one embodiment, the coupled model includes a WRF-Fire model comprising an atmospheric module and a fire behavior module.

[0051] In one embodiment, the adjustable parameters of the atmospheric physics module include at least the turbulent mixing length or the TKE closure constant in the planetary boundary layer (PBL) parameterization scheme.

[0052] In one embodiment, the adjustable parameters of the surface physics module include at least the equilibrium moisture content deviation term of the combustible material moisture content time-delay model.

[0053] Secondly, this application also provides a forest fire situation awareness method based on optical active and passive remote sensing data, including:

[0054] S100, construct an augmented state vector in memory, and run the coupled model based on the augmented state vector to generate model predictions; wherein, the augmented state vector contains the dynamic state variables of the coupled model, the adjustable physical parameters of at least one atmospheric physics module, and the adjustable physical parameters of at least one surface physics module; the coupled model contains at least one atmospheric physics module and one surface physics module.

[0055] S200: Acquire real-time atmospheric observation data and real-time surface observation data; calculate the residuals between the real-time atmospheric observation data and real-time surface observation data and the predicted values; and update the augmented state vector based on the residuals using a data assimilation algorithm. The updated augmented state vector includes the updated adjustable physical parameters and the corresponding parameter change trend terms.

[0056] S300 generates forest fire situational awareness products based on the updated augmented state vector running coupled model.

[0057] In one embodiment, generating a forest fire situation awareness product by running a coupled model based on the updated augmented state vector includes: extracting the updated adjustable physical parameters and corresponding parameter change trend terms from the updated augmented state vector; constructing a dynamic parameter prediction model based on the updated adjustable physical parameters and corresponding parameter change trend terms; injecting the dynamic parameter prediction model into the coupled model to generate an adaptively optimized coupled model; and generating a forest fire situation awareness product based on the output of the adaptively optimized coupled model.

[0058] In one embodiment, a computer device is provided, which may be a server. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores relevant data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a forest fire situational awareness method based on optical active and passive remote sensing data.

[0059] In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, memory, a communication interface, a display screen, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a forest fire situational awareness method based on optical active and passive remote sensing data.

[0060] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0061] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0062] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0063] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0064] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0065] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] The embodiments described above are merely examples of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

Claims

1. A forest fire situation awareness system based on optical active and passive remote sensing data, characterized in that, include: An optical active remote sensing subsystem includes at least one Doppler wind lidar for acquiring real-time atmospheric observation data over the target area; Real-time atmospheric observation data includes turbulent kinetic energy profiles or vertical velocity variance; The optical passive remote sensing subsystem, including a satellite remote sensing platform, is used to acquire real-time surface observation data of the target area; Real-time surface observation data includes calculations of real-time fire line locations using thermal infrared and shortwave infrared bands; The ground processing and modeling unit includes a processor and memory, and the processor is configured to perform the following steps: An augmented state vector is constructed in memory, and the ensemble of the coupled model is run based on the augmented state vector to generate model predictions. The augmented state vector includes the dynamic state variables of the coupled model, adjustable physical parameters of at least one atmospheric physics module, adjustable physical parameters of at least one surface physics module, and a parameter variation trend term for at least one of the adjustable physical parameters. The coupled model includes at least one atmospheric physics module and one surface physics module, including a WRF-Fire model containing an atmospheric module and a fire behavior module. The adjustable parameters of the atmospheric physics module include at least the turbulent mixing length or TKE closure constant in the planetary boundary layer parameterization scheme. The parameter variation trend term corresponding to the adjustable parameters of the atmospheric physics module includes the time rate of change of the turbulent mixing length. The adjustable parameters of the surface physics module include at least the equilibrium water content deviation term of the combustible water content time-delay model. The parameter variation trend term corresponding to the adjustable parameters of the surface physics module includes the time rate of change of the equilibrium water content deviation term. Calculate the residuals between real-time atmospheric observation data, real-time surface observation data and predicted values, and update the augmented state vector based on the residuals using a data assimilation algorithm; Forest fire situation awareness products are generated by running a coupled model based on the updated augmented state vector.

2. A forest fire situation awareness method based on optical active and passive remote sensing data, characterized in that, include: An augmented state vector is constructed in memory, and a coupled model is run based on the augmented state vector to generate model predictions. The augmented state vector includes dynamic state variables of the coupled model, adjustable physical parameters of at least one atmospheric physics module, adjustable physical parameters of at least one surface physics module, and a parameter variation trend term for at least one of the adjustable physical parameters. The coupled model includes at least one atmospheric physics module and one surface physics module, including a WRF-Fire model comprising an atmospheric module and a fire behavior module. The adjustable parameters of the atmospheric physics module include at least the turbulent mixing length or TKE closure constant in the planetary boundary layer parameterization scheme. The parameter variation trend term corresponding to the adjustable parameters of the atmospheric physics module includes the time rate of change of the turbulent mixing length. The adjustable parameters of the surface physics module include at least the equilibrium water content deviation term of the combustible water content time-delay model. The parameter variation trend term corresponding to the adjustable parameters of the surface physics module includes the time rate of change of the equilibrium water content deviation term. Real-time atmospheric observation data and real-time surface observation data are acquired. The residuals between the real-time atmospheric observation data, real-time surface observation data and predicted values ​​are calculated. The augmented state vector is updated based on the residuals using a data assimilation algorithm. The updated augmented state vector includes the updated adjustable physical parameters and the corresponding parameter change trend terms. The real-time atmospheric observation data includes turbulent kinetic energy profiles or vertical velocity variances. The real-time surface observation data includes the real-time fire line location calculated using thermal infrared and shortwave infrared bands. Forest fire situation awareness products are generated by running a coupled model based on the updated augmented state vector.

3. The method according to claim 2, characterized in that, Forest fire situation awareness products are generated by running the coupled model based on the updated augmented state vector, including: Extract the updated adjustable physical parameters and corresponding parameter change trend terms from the augmented state vector after the update; A dynamic parameter prediction model is constructed based on the updated adjustable physical parameters and the corresponding parameter change trend terms. The dynamic parameter model is injected into the coupled model to generate an adaptively optimized coupled model, and the output of the adaptively optimized coupled model is used to generate a forest fire situation awareness product.

Images