Forest fires, severe natural disasters causing substantial damage, necessitate accurate predictive modeling to guide preventative measures effectively. This study introduces an enhanced window-based Transformer time series forecasting model aimed at improving the precision of forest fire predictions. Leveraging time series data from 2020 to 2021 in Chongli, a myriad of forest fire influencing factors were ascertained using remote sensing satellite and GIS technologies, with their interrelationships estimated through a multicollinearity test. Given the intricate nature of real-world forest fire prediction tasks, we propose a novel window-based Transformer architecture complemented by a dual time series input strategy premised on 13 influential factors. Subsequently, time series data were incorporated into the model to generate a forest fire risk prediction map in Chongli District. The model’s effectiveness was then evaluated using various metrics, including accuracy (ACC), root mean square error (RMSE), and mean absolute error (MAE), and compared with traditional deep learning methods. Our model demonstrated superior predictive performance (ACC = 91.56%, RMSE = 0.37, MAE = 0.05), harnessing spatial background information efficiently and effectively utilizing the periodicity of forest fire factors. Consequently, the study proves this method to be a novel and potent approach for time series fire prediction.
Geo-Kompsat-2A (GK-2A) is the third new-generation geostationary meteorological satellite that orbits Asia and monitors China and its surrounding areas, following the Himawari-8 and Fengyun-4A satellites. The nadir point positioning and satellite channel parameters of the GK-2A are better than those of the Himawari-8 and FY-4A, which are more conducive to fire monitoring in China. In this study, a new fire detection algorithm is proposed based on GK-2A satellite data. That is, considering the large solar zenith angle correction for reflectance and the proportion information of background pixels in the existing spatial threshold method, fires under the different underlying surface types and solar radiation states can be automatically identified. Moreover, the accuracy of the Himawari-8 fire monitoring algorithm and the present algorithm of GK-2A is compared and analyzed through the ground truth fire spot data. The results show that compared with the original fire monitoring algorithm with fixed parameter thresholds, the brightness temperature difference of this algorithm is reduced by 0.55 K, and the correction coefficient is reduced by 0.6 times, the fire can be found earlier, and the monitoring sensitivity is improved. According to the practical fire case, the present fire monitoring algorithm of GK-2A has better monitoring accuracy than the fire monitoring algorithm of Himawari-8. The present fire monitoring algorithm of GK-2A can meet the fire monitoring requirements under different sun angles, different cloud cover ratios and vegetation ratios with good versatility.
In general, the far-infrared channel in the wavelength range of 10.5–12.0 µm plays an auxiliary role in wildfire detection as its sensitivity to high-temperature targets is far lower than the mid-infrared channel in the wavelength range of 3.5–4.0 µm at the same spatial resolution (1 km, which is the spatial resolution of infrared channels in most satellites used for wildfire monitoring in daily operational mode). The Medium-Resolution Spectral Imager II onboard the Fengyun-3D polar orbiting meteorological satellite (FY-3D/MERSI-II) contains far-infrared channels with a spatial resolution of 250 m at the wavelengths of 10.8 μm and 12.0 μm, which promotes the application of far-infrared channels in wildfire monitoring. In this study, the features of FY-3D/MERSI-II far-infrared channels in fire monitoring are discussed. The sensitivity of 10.8 μm (250 m) to fire spots and the influence of solar radiation reflection on the infrared channels are quantitatively analyzed. The method of using 10.8 μm (250 m) as a major data source to detect fire spots is proposed, and several typical wildfire cases are used to verify the proposed method. The results show that the 10.8 μm (250 m) far-infrared channel has the same advantages as the existing method in wildfire monitoring in terms of a more precise positioning of the detected fire pixel, avoiding interference by solar radiation reflections, and reflecting stronger fire regions in large fire fields.
18 June 2022 in International Journal of Remote Sensing
Fengyun-3E (FY-3E) is the world’s first operational meteorological satellite on dawn-dusk orbit. In order to realize the generality of algorithms for different satellites and provide a basis for subsequent inversion studies on remote sensing parameters, such as land surface temperature, fire, and atmospheric water vapour, we perform a cross-comparison of the brightness temperature (TB) of the six infrared (IR) channels between the Medium Resolution Spectral Imager Level L onboard the FY-3E satellite (FY-3E/MERSI-LL) and the Advanced Himawari Imager onboard the Himawari-8 satellite (Himawari-8/AHI). The statistical indicators, correlation coefficient, root mean square error (RMSE) and mean bias, are used to analyse the consistency and differences of the channel parameters between the MERSI-LL and AHI in terms of overall data, different regions and different TB. The results show that the TB data observed by the FY-3E/MERSI-LL and Himawari-8/AHI are generally consistent, with correlation coefficients of more than 0.99, mean biases of less than 1.60 K and RMSEs of less than 1.90 K. In terms of different channels, the TB in channel IR 038 has the best consistency between the two sensors, while the results from the MERSI-LL are higher than those of the AHI for channels IR038, IR086, IR108 and IR120, and lower than those of the AHI for channel IR041. Analysing the situation in different regions, we find that the MERSI-LL and AHI data are highly consistent and relatively stable in the Huang-Huai Region, with low RMSEs, low mean biases and high correlation coefficients (more than 0.95) at all channels. For the different TB ranges, the RMSEs and mean biases for the TB ranges of 270–280 K and 280–290 K between the MERSI-LL and AHI are small, and the data in the 290–300 K TB range has great differences and fluctuations between the two sensors.
In this study, we compare the data of the advanced geostationary radiation imager (AGRI) on board the FY-4B and the advanced meteorological imager (AMI) on board the GK-2A, in terms of overall data, different reflectivity/brightness temperature intervals, different regions, and different underlying surfaces. The results show that the AGRI and AMI data are generally consistent; the mean biases for reflectivity channels show a range of 0.50% to 1.69%, with channel VIR004 being exceptionally good, while brightness temperature (TB) differences in the IR channels ranging from 0.11 to 0.57 K, with channel IR120 being the most accurate. The reflectivity of the AGRI is higher than that of the AMI in terms of mean bias. The dispersion of the reflectivity difference between the AGRI and AMI is smaller at the short-wavelength channels than that at the longer-wavelength channels. The TB data observed by the AGRI are higher than those of AMI at conditions above 310 K. In the case of observing the same target, the difference in infrared brightness temperature due to the random noise signal is small. The differences between the two sensors can be considerably reduced by revising mean biases. In the following studies of quantitative product algorithms, the characteristics of sensor data need to be further analyzed in detail.
This study presents a GIS-based method integrating hourly transport pathways and wind-field grid reconstruction, straw open burning (SOB) source identification, and a two-stage spatiotemporal multi-box modeling approach to quantify the contribution of external sources of SOB to elevated urban PM2.5 concentrations during a specific pollution episode (PE) at a high temporal resolution of 1 h. Taking Jilin Province as an empirical study, the contribution of SOB in province-wide farmlands to urban haze episodes in Changchun during the SOB season of 2020–2021 was evaluated quantitatively using a combination of multi-source datasets. The results showed that Changchun experienced three severe PEs and one heavy PE during the study period, and the total PM2.5 contributions from SOB sources were 352 μg m−3, 872 μg m−3, and 1224 μg m−3 during the three severe PEs, respectively; these accounted for 7%, 27%, and 23% of the urban cumulative PM2.5 levels, which were more obvious than the contribution during the PE. The total PM2.5 contribution from SOB sources (4.9 μg m−3) was only 0.31% of the urban cumulative PM2.5 level during the heavy PE. According to the analysis of the impact of individual factors, some policy suggestions are put forward for refined SOB management, including control spatial scope, burning time interval, as well as burning area limit under different urban and transport pathways’ meteorological conditions and different transport distances.