Detection and Monitoring of Forest Fires Using Himawari-8 Geostationary Satellite Data in South Korea

Jang, Eunna; Kang, Yoojin; Im, Jungho; Lee, Dongwon; Yoon, Jongmin; Kim, Sang-Kyun

doi:10.3390/rs11030271

Cited by 84 publications

(36 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…All decision trees are aggregated using (weighted) majority voting for classification and (weighted) averaging for regression. RF provides the relative importance of independent variables, which has been widely used in previous studies, even though it is of local importance, not global [14,41,[69][70][71][72]. It can be obtained as a percentage of the increased MSE for each variable using out-of-bag (OOB) data.…”

Section: Step 2: Random Forest (Rf)mentioning

confidence: 99%

Short-Term Forecasting of Satellite-Based Drought Indices Using Their Temporal Patterns and Numerical Model Output

Park

Han

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Drought forecasting is essential for effectively managing drought-related damage and providing relevant drought information to decision-makers so they can make appropriate decisions in response to drought. Although there have been great efforts in drought-forecasting research, drought forecasting on a short-term scale (up to two weeks) is still difficult. In this research, drought-forecasting models on a short-term scale (8 days) were developed considering the temporal patterns of satellite-based drought indices and numerical model outputs through the synergistic use of convolutional long short term memory (ConvLSTM) and random forest (RF) approaches over a part of East Asia. Two widely used drought indices—Scaled Drought Condition Index (SDCI) and Standardized Precipitation Index (SPI)—were used as target variables. Through the combination of temporal patterns and the upcoming weather conditions (numerical model outputs), the overall performances of drought-forecasting models (ConvLSTM and RF combined) produced competitive results in terms of r (0.90 and 0.93 for validation SDCI and SPI, respectively) and nRMSE (0.11 and 0.08 for validation of SDCI and SPI, respectively). Furthermore, our short-term drought-forecasting model can be effective regardless of drought intensification or alleviation. The proposed drought-forecasting model can be operationally used, providing useful information on upcoming drought conditions with high resolution (0.05°).

show abstract

Section: Step 2: Random Forest (Rf)mentioning

confidence: 99%

Short-Term Forecasting of Satellite-Based Drought Indices Using Their Temporal Patterns and Numerical Model Output

Park

Han

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Regarding wildfire safety investigations satellite-based measurement technology can be applied to assist in determining causes of an event, conditions affecting the severity of the outcome in terms of damage to ecosystems and human society as well as potential remediation. Furthermore, this would allow the measurement and analysis of wildfire-caused emissions (Li et al, 2000;Jang et al, 2019;Hislop et al, 2020). The satellite-based emission measurements combined with fire activity and vegetation productivity have been successfully applied internationally in the Global Fire Emission Database (2021).…”

Section: Comparison Of Technical Aspects Between International Aircraft Accidents and Proposed International Wildfire Investigationsmentioning

confidence: 99%

Wildfire-Related Catastrophes: The Need for a Modern International Safety Investigation Procedure

Vuorio¹,

Kovanen²,

Budowle³

et al. 2021

Front. Clim.

View full text Add to dashboard Cite

Despite the increased frequency and scale of wildfire-related catastrophes, there has been little or no effective and coordinated international policy to address their highly negative impact. Possibly a generalized approach to respond to such major events could be modeled on existing international safety investigation policies and agreements that already have proved successful. The International Civil Aviation Organization (ICAO) outlines safety investigations after international fatal aviation accidents. Although this well-established safety investigation protocol cannot be directly applied in acute wildfire-related accidents, it can offer a useful framework for establishing international guidelines to reduce risk of future wildfire catastrophes. The co-operation between safety investigation authorities has been shown to be fruitful especially for those less developed countries that have limited resources and experience related to accident investigations. While primarily an adaptive measure that can set practices to reduce vulnerability and fragility of ecosystems and human societies, the same policies could be seen as a climate change mitigation measure, as wildfires can contribute significantly to global CO2 emissions. Finally, the concept of independent and qualified safety investigations represents the principle of serendipity: disclosing by accident something that has not been foreseen. Feedback from reality compensates assumptions and limitations of feedforward analysis of complex systems that can only reveal their dynamics and performance in reality and over time.

show abstract

“…Hotspot detection algorithms and the science of fire detection have been widely discussed by Lee and Tag (1990), Justice et al, 1993, Flasse andCeccato (1996), Setzer and Malingreau (1996), Chu et al, 1998), Giglio et al, (1999Giglio et al, ( , 2003Giglio et al, ( , 2008Giglio et al, ( , 2009, Giglo (2007 and 2009), Kushida, (2006), Li et al, (2001), Morisette et al, (2002), Rogers et al, (2006), Molina-Pico et al, (2016), Lutakamale and Kaijage, (2017), Mubarak et al, (2018) and Jang et al, (2019), Gibson, et al, (2020), andFIRMS, (2021). Most hotspot algorithms use fixed thresholds applied to single or multiple bands of thermal IR data for both MODIS and AVHRR sensors and contextual analysis of the pixel and its background (Li et al, 2001).…”

Section: Relevant Studiesmentioning

confidence: 99%

Validation of MODIS and AVHRR Fire Detections in Australia

Forghani

Thankappan

Cechet

2021

IJG

View full text Add to dashboard Cite

The Sentinel Bushfire Monitoring System is an internet-based mapping tool which provides timely spatial information to fire agencies across Australia. The mapping system allows users to identify active fire locations that pose a potential risk to communities and property. Sentinel at Geoscience Australia currently provides hotspot information derived from Moderate Resolution Imaging Spectroradiometer (MODIS) and Advanced Very High Resolution Radiometer (AVHRR) sensors on a continent-wide and daily basis enabling the fire community and general public to locate active fires. There has been little validation undertaken of the Sentinel since the system began operating in November 2002. Validation datasets have been collected for this work during the 2003-2007 fire seasons. Five study areas were selected to validate the detection capabilities of the MODIS and AVHRR hotspot product with fire activity that was mapped using high resolution earth observation imagery. The objective is to evaluate the reliability with which hotspots identified in MODIS and AVHRR thermal data can be used to identify fires. This consists of comparing the accuracy of AVHRR versus MODIS and quantifying the accuracy of both products. This objective is achieved by characterising errors through a stratified random sampling technique establishing a relationship between the ‘fire’ and ‘no fire’ condition, and error assessment using multi- source reference datasets over coincident MODIS and AVHRR pixels. The validation framework comprised two key approaches including validation of AVHRR hotspots in relation to MODIS hotspots and validation of both MODIS and AVHRR hotspots using multi-sensor earth observation imagery datasets. The study identified sources of errors associated with the Sentinel hotspots which could be used to improve the performance of hotspot algorithms and provide user-friendly information for the users. Statistical analysis revealed that overall commission errors of MODIS and AVHRR hotspots over the 5% sample data were 15% and 68% respectively, and overall omission errors of MODIS and AVHRR hotspots were 17% and 23% respectively. An important outcome of this study is the production of a database of fire locations derived from high-resolution imagery, which can serve as a resource for future validation efforts as detection algorithms evolve and sensors change.

show abstract

Detection and Monitoring of Forest Fires Using Himawari-8 Geostationary Satellite Data in South Korea

Cited by 84 publications

References 45 publications

Short-Term Forecasting of Satellite-Based Drought Indices Using Their Temporal Patterns and Numerical Model Output

Short-Term Forecasting of Satellite-Based Drought Indices Using Their Temporal Patterns and Numerical Model Output

Wildfire-Related Catastrophes: The Need for a Modern International Safety Investigation Procedure

Validation of MODIS and AVHRR Fire Detections in Australia

Contact Info

Product

Resources

About