the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Meteorological Analysis of the Forcett-Dunalley Wildfire in 2013 in Tasmania, Australia
Abstract. A major bushfire occurred during January 2013 near the towns Forcett and Dunalley in southeast Tasmania, Australia. Several records were broken by this wildfire, in terms of impacts to eco-systems, infrastructure and lives, as well as the first documented fire storm development in Tasmania in the form of pyrocumulonimbus. The Australian Bureau of Meteorology high-resolution regional reanalysis for Tasmania (BARRA-TA), with 1.5-km spatial resolution, together with in-situ observations, was applied to reconstruct the wildfire event. The antecedent climatic conditions in Tasmania included large increase in fuel load due to abundant rain one to two years before the event, followed by a heatwave during the summer of 2012/13. In the three periods we identified during the event reconstruction, the second period was the most dramatic, in which a low-level jet was directed downslope in southeast Tasmania to accelerate the fire spread. Moreover, spotting of over 3 km was observed, and pyrocumulonimbus developed in this period with lightning up to 13 km from the fire. A cold front crossed the fireground during the third period, and thus played a different role compared with some past extreme fire events in terms of lifting and wind direction change. Our analyses conclude that climatic conditions, synoptic patterns and mesoscale convective environment all contributed to this wildfire event.
- Preprint
(15330 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on nhess-2023-210', Anonymous Referee #1, 04 Mar 2024
This is a case study describing some local weather observations and reanalysis data around the time of an extreme wildfire event in Tasmania. It is an interesting event, noting several studies publishing on this previously (including by some of these same authors https://www.mdpi.com/2073-4433/14/7/1076 and https://doi.org/10.5194/nhess-2019-354).
Some things could be enhanced in this study, including more supporting results for conclusions made. Examples for that include the following statements in the Abstract:
- The first statement in the Abstract after the data description says "antecedent climatic conditions in Tasmania included large increase in fuel load due to abundant rain one to two years before the event". But it is not shown in the results if there actually was a large increase in fuel load compared to normal. Details on what drives vegetation growth in that region are also not presented, such as if factors other than rain like soil temperature could potentially be relevant too.
- The next sentence in the Abstract says "a low-level jet was directed downslope in southeast Tasmania to accelerate the fire spread" but it is hard to determine from the results presented that the jet caused accelerated fire spread.
- The next sentence in the Abstract says "spotting of over 3 km was observed, and pyrocumulonimbus developed in this period with lightning up to 13 km from the fire". Spotting data and lightning data are not presented, or analysis to demonstrate pyrocumulonimbus occurrence (distinct from pyrocumulus).
- The final sentence in the Abstract says "Our analyses conclude that climatic conditions, synoptic patterns and mesoscale convective environment all contributed to this wildfire event", while noting previous studies have also discussed those types of aspects including the studies mentioned above on this event by some of these authors.
In the data section of the manuscript, a key part of this is describing the FFDI and data used to calculate it. None of the figures show results for FFDI, but it could have been interesting to see examples leading up to and during FFDI being above 100, given that is discussed several times in the study (for example the 1-minute AWS data used in this study to calculate FFDI seems could have shown how variable the values are around their peaks). In various parts of the study the use of specific humidity, rather than relative humidity, would avoid ambiguity when examining moisture variability (given the temperature component of relative humidity).
In general throughout, the study could be enhanced for accuracy. Examples for the section of text from lines 457-486:
- In this section of text it says "conditions supporting pyroCb or firestorm development are similar to those generating conventional thunderstorms (Tory and Kepert, 2021). The unstable atmospheric conditions are essential for pyrocumulus cloud formation". However, a key defining aspect of pyroCb occurrence is the influence of the fire on atmospheric conditions, which can sometimes occur in atmospheric conditions that are moderately or marginally stable, in contrast to atmospheric conditions in which conventional thunderstorms typically form.
- This section of text also seems to suggest the relative humidity around 500 hPa helped trigger pyroCb formation, but it is not clear how that occurred from the results presented (similarly also for lines 527-528).
- The text here also says "fire intensity, which refers to the amount of energy generated" however, intensity here is (presumably) a rate of energy generated that might be integrated over time to give amount of energy.
- "Firestorm" is used here, and many times through the manuscript (or "fire storm" in some places). This could be kept consistent and clarified in definition for what it refers, as sometimes it seems to refer to fire-generated thunderstorms but not in other cases (such as lines 485-486 referring to a 1991 study not on pyroCbs).
Citation: https://doi.org/10.5194/nhess-2023-210-RC1 - AC1: 'Reply on RC1', kevin cheung, 22 Apr 2024
-
RC2: 'Comment on nhess-2023-210', Anonymous Referee #2, 09 Mar 2024
This study presents a meteorological analysis of the 2013 Forcett-Dunalley bushfire. The subject matter is within the scope of the NHESS journal, and both the title and the abstract accurately describe the contents of the study. Although the analysis from this study is applicable and useful, it is not completely clear if it represents a specific advance in knowledge. Other studies that analyze the fire and meteorological conditions for 2013 Forcett-Dunalley bushfire have already been published. The abstract and introduction sections can be revised to make it easier for the reader to identify which are the specific advances in knowledge offered by this study. Specific comments are listed below:
- Line 23: The abstract states that spotting of over 3 km was observed, however subsequent sections reduce this distance to 2.5 km (lines 98 and 136).
- Lines 67-68: The terminology is confusing. Technically speaking, a fire burns until it is extinguished. If the fire was not extinguished before March 20, it did not stop burning on January 18. Please use more specific terminology to describe what conditions changed for the fire on that date (e.g., was it controlled?)
- Lines 149-150: It is stated that ‘the selected meteorological stations included Hobart […] and Hobart Airport’. Were these two stations the only ones that were considered for the analysis? If yes, why were the other stations shown in Fig. 1a not included (specially Dunalley)? If no, an improvement in the writing of this section can be beneficial for the readers.
- Line 136: The information provided does not explain how it was determined that the bushfire in the Tasman Peninsula was initiated by fire spotting originated in Eagle Neck, nor how the 2.5 km distance (or 3 km according to the abstract) was obtained. Moreover, while discussing fire spotting the readers are referred to Fig. 1a, but this figure focuses exclusively on presenting the location of weather stations and does not clearly identify in which location of the Tasman Peninsula the spotted bushfire was ignited.
- Line 175: The last two rows of Table 2 list ‘???’ as the description of the difficulty of fire suppression of events during extreme and catastrophic FFDR. It is clear from the description for severe FFDR that suppression cannot be expected at that level of danger level and above, however it would be convenient to provide descriptions for the last two categories (e.g., fire suppression is considered impossible).
- Line 210: It is stated that the rainfall in 2009 and 2011 was ‘very much above average’, but no quantitative values are provided nor is it clear what the frequency and period over which the average was calculated are. Here (and previously in lines 207-208) the term ‘very much’ can be interpreted as subjective if no additional details are provided (e.g., a graph showing the yearly rainfall can provide important context for the readers).
Technical corrections are listed below:
- Lines 42-43: Percentages do not add up to 100%
- Line 395-426: These paragraphs repeat information that has already been presented in preceding sections of the study. This information can be omitted or summarized without compromising the message of the paper.
- Line 470: ‘low level jet’ can be replaced with ‘LLJ’ to make it consisted with the rest of the document.
Citation: https://doi.org/10.5194/nhess-2023-210-RC2 - AC2: 'Reply on RC2', kevin cheung, 22 Apr 2024
-
RC3: 'Comment on nhess-2023-210', Anonymous Referee #3, 09 Mar 2024
This study aims to reconstruct the meteorological conditions leading up to and occurring during the Forcett-Dunalley wildfire. Although the subject matter seems appropriate for NHESS, I have some questions about the advances this work is making over previously published work and BoM reports. The work collects in one place, the different contributing weather factors to the fire and pyroCb storm, but I’m a bit unclear what original analysis the authors contributed here.
The study uses mostly data from the BoM and the BARRA reanalysis, but there are places throughout the paper where more clarification is needed on the source of the observations. For example, in sec 2.1, line 98: how was flame height measured and how was spotting distance reported? Were these spot fires someone observed or reports of embers falling 2.5 km away from the front? And where are the numbers for spread rate coming from? Did the authors measure this themselves from satellite images or some other method? Or are those numbers from the BoM 2013 report? Is there any ability to quantify uncertainty on these measurements?
Similarly for sec 2.2: what are the sources for some of these numbers? Are the spread rates coming from the BoM reports, or Ndalila 2018 or from analysis the authors did themselves? Reporting that “Upon arrival in Dunalley, the fire spread rate was 45 to 50m min-1” makes it unclear what information the authors are collecting from other sources and what they are calculating themselves.
Occasionally the authors make claims without quantifying anything specific to support those claims. For example, when talking about the wetter years in 2009 and 2011 that led to more dense vegetation, increasing the fuel load. These claims are plausible, but the authors should provide support since fuel load is a key parameter for wildfire growth. How much more rainfall was there in those years and can they quantify the fuel load changes over the years before the fire? Without backup, this section reads more as speculation, but could be strengthened greatly with some quantification since estimating fuel load would also be valuable to modelers. Even in the FFDI calculation, is the assumption of a fixed value of 12.5 tons per hectare of fuel load appropriate given what was stated earlier about the denser vegetation?
I also feel the paper could benefit from a reorganization, especially in the results & discussion section. The discussion section was very clear and helped contextualize the earlier portions of the results (which are very detailed play-by-play style reporting). I think intertwining the descriptive parts of the results with their relative discussion could make the importance of those observations much more clear and immediate.
The paper could also use some language editing for clarity and grammar, and working hyperlinks for citations and figures and tables would be helpful.
Finally, the figures in this paper need to be either vectorized (where appropriate) or much higher resolution. Some of the labeling in the figures or the figures themselves can be hard to read even when zooming in. Color maps used could also be improved for clarity, for example in figure 8, since each line represents an evolution in time, a perceptually-uniform sequential colormap could be used to make the figure more readable, even in black & white. Other figures could also benefit from the use of perceptually uniform colormaps, instead of the default jet. Figures 9, 10, 11, 12, 13, 14, 15. Such colormaps not only improve readability but also make figures accessible to those with different forms of color vision deficiency.
Citation: https://doi.org/10.5194/nhess-2023-210-RC3 - AC3: 'Reply on RC3', kevin cheung, 22 Apr 2024
Status: closed
-
RC1: 'Comment on nhess-2023-210', Anonymous Referee #1, 04 Mar 2024
This is a case study describing some local weather observations and reanalysis data around the time of an extreme wildfire event in Tasmania. It is an interesting event, noting several studies publishing on this previously (including by some of these same authors https://www.mdpi.com/2073-4433/14/7/1076 and https://doi.org/10.5194/nhess-2019-354).
Some things could be enhanced in this study, including more supporting results for conclusions made. Examples for that include the following statements in the Abstract:
- The first statement in the Abstract after the data description says "antecedent climatic conditions in Tasmania included large increase in fuel load due to abundant rain one to two years before the event". But it is not shown in the results if there actually was a large increase in fuel load compared to normal. Details on what drives vegetation growth in that region are also not presented, such as if factors other than rain like soil temperature could potentially be relevant too.
- The next sentence in the Abstract says "a low-level jet was directed downslope in southeast Tasmania to accelerate the fire spread" but it is hard to determine from the results presented that the jet caused accelerated fire spread.
- The next sentence in the Abstract says "spotting of over 3 km was observed, and pyrocumulonimbus developed in this period with lightning up to 13 km from the fire". Spotting data and lightning data are not presented, or analysis to demonstrate pyrocumulonimbus occurrence (distinct from pyrocumulus).
- The final sentence in the Abstract says "Our analyses conclude that climatic conditions, synoptic patterns and mesoscale convective environment all contributed to this wildfire event", while noting previous studies have also discussed those types of aspects including the studies mentioned above on this event by some of these authors.
In the data section of the manuscript, a key part of this is describing the FFDI and data used to calculate it. None of the figures show results for FFDI, but it could have been interesting to see examples leading up to and during FFDI being above 100, given that is discussed several times in the study (for example the 1-minute AWS data used in this study to calculate FFDI seems could have shown how variable the values are around their peaks). In various parts of the study the use of specific humidity, rather than relative humidity, would avoid ambiguity when examining moisture variability (given the temperature component of relative humidity).
In general throughout, the study could be enhanced for accuracy. Examples for the section of text from lines 457-486:
- In this section of text it says "conditions supporting pyroCb or firestorm development are similar to those generating conventional thunderstorms (Tory and Kepert, 2021). The unstable atmospheric conditions are essential for pyrocumulus cloud formation". However, a key defining aspect of pyroCb occurrence is the influence of the fire on atmospheric conditions, which can sometimes occur in atmospheric conditions that are moderately or marginally stable, in contrast to atmospheric conditions in which conventional thunderstorms typically form.
- This section of text also seems to suggest the relative humidity around 500 hPa helped trigger pyroCb formation, but it is not clear how that occurred from the results presented (similarly also for lines 527-528).
- The text here also says "fire intensity, which refers to the amount of energy generated" however, intensity here is (presumably) a rate of energy generated that might be integrated over time to give amount of energy.
- "Firestorm" is used here, and many times through the manuscript (or "fire storm" in some places). This could be kept consistent and clarified in definition for what it refers, as sometimes it seems to refer to fire-generated thunderstorms but not in other cases (such as lines 485-486 referring to a 1991 study not on pyroCbs).
Citation: https://doi.org/10.5194/nhess-2023-210-RC1 - AC1: 'Reply on RC1', kevin cheung, 22 Apr 2024
-
RC2: 'Comment on nhess-2023-210', Anonymous Referee #2, 09 Mar 2024
This study presents a meteorological analysis of the 2013 Forcett-Dunalley bushfire. The subject matter is within the scope of the NHESS journal, and both the title and the abstract accurately describe the contents of the study. Although the analysis from this study is applicable and useful, it is not completely clear if it represents a specific advance in knowledge. Other studies that analyze the fire and meteorological conditions for 2013 Forcett-Dunalley bushfire have already been published. The abstract and introduction sections can be revised to make it easier for the reader to identify which are the specific advances in knowledge offered by this study. Specific comments are listed below:
- Line 23: The abstract states that spotting of over 3 km was observed, however subsequent sections reduce this distance to 2.5 km (lines 98 and 136).
- Lines 67-68: The terminology is confusing. Technically speaking, a fire burns until it is extinguished. If the fire was not extinguished before March 20, it did not stop burning on January 18. Please use more specific terminology to describe what conditions changed for the fire on that date (e.g., was it controlled?)
- Lines 149-150: It is stated that ‘the selected meteorological stations included Hobart […] and Hobart Airport’. Were these two stations the only ones that were considered for the analysis? If yes, why were the other stations shown in Fig. 1a not included (specially Dunalley)? If no, an improvement in the writing of this section can be beneficial for the readers.
- Line 136: The information provided does not explain how it was determined that the bushfire in the Tasman Peninsula was initiated by fire spotting originated in Eagle Neck, nor how the 2.5 km distance (or 3 km according to the abstract) was obtained. Moreover, while discussing fire spotting the readers are referred to Fig. 1a, but this figure focuses exclusively on presenting the location of weather stations and does not clearly identify in which location of the Tasman Peninsula the spotted bushfire was ignited.
- Line 175: The last two rows of Table 2 list ‘???’ as the description of the difficulty of fire suppression of events during extreme and catastrophic FFDR. It is clear from the description for severe FFDR that suppression cannot be expected at that level of danger level and above, however it would be convenient to provide descriptions for the last two categories (e.g., fire suppression is considered impossible).
- Line 210: It is stated that the rainfall in 2009 and 2011 was ‘very much above average’, but no quantitative values are provided nor is it clear what the frequency and period over which the average was calculated are. Here (and previously in lines 207-208) the term ‘very much’ can be interpreted as subjective if no additional details are provided (e.g., a graph showing the yearly rainfall can provide important context for the readers).
Technical corrections are listed below:
- Lines 42-43: Percentages do not add up to 100%
- Line 395-426: These paragraphs repeat information that has already been presented in preceding sections of the study. This information can be omitted or summarized without compromising the message of the paper.
- Line 470: ‘low level jet’ can be replaced with ‘LLJ’ to make it consisted with the rest of the document.
Citation: https://doi.org/10.5194/nhess-2023-210-RC2 - AC2: 'Reply on RC2', kevin cheung, 22 Apr 2024
-
RC3: 'Comment on nhess-2023-210', Anonymous Referee #3, 09 Mar 2024
This study aims to reconstruct the meteorological conditions leading up to and occurring during the Forcett-Dunalley wildfire. Although the subject matter seems appropriate for NHESS, I have some questions about the advances this work is making over previously published work and BoM reports. The work collects in one place, the different contributing weather factors to the fire and pyroCb storm, but I’m a bit unclear what original analysis the authors contributed here.
The study uses mostly data from the BoM and the BARRA reanalysis, but there are places throughout the paper where more clarification is needed on the source of the observations. For example, in sec 2.1, line 98: how was flame height measured and how was spotting distance reported? Were these spot fires someone observed or reports of embers falling 2.5 km away from the front? And where are the numbers for spread rate coming from? Did the authors measure this themselves from satellite images or some other method? Or are those numbers from the BoM 2013 report? Is there any ability to quantify uncertainty on these measurements?
Similarly for sec 2.2: what are the sources for some of these numbers? Are the spread rates coming from the BoM reports, or Ndalila 2018 or from analysis the authors did themselves? Reporting that “Upon arrival in Dunalley, the fire spread rate was 45 to 50m min-1” makes it unclear what information the authors are collecting from other sources and what they are calculating themselves.
Occasionally the authors make claims without quantifying anything specific to support those claims. For example, when talking about the wetter years in 2009 and 2011 that led to more dense vegetation, increasing the fuel load. These claims are plausible, but the authors should provide support since fuel load is a key parameter for wildfire growth. How much more rainfall was there in those years and can they quantify the fuel load changes over the years before the fire? Without backup, this section reads more as speculation, but could be strengthened greatly with some quantification since estimating fuel load would also be valuable to modelers. Even in the FFDI calculation, is the assumption of a fixed value of 12.5 tons per hectare of fuel load appropriate given what was stated earlier about the denser vegetation?
I also feel the paper could benefit from a reorganization, especially in the results & discussion section. The discussion section was very clear and helped contextualize the earlier portions of the results (which are very detailed play-by-play style reporting). I think intertwining the descriptive parts of the results with their relative discussion could make the importance of those observations much more clear and immediate.
The paper could also use some language editing for clarity and grammar, and working hyperlinks for citations and figures and tables would be helpful.
Finally, the figures in this paper need to be either vectorized (where appropriate) or much higher resolution. Some of the labeling in the figures or the figures themselves can be hard to read even when zooming in. Color maps used could also be improved for clarity, for example in figure 8, since each line represents an evolution in time, a perceptually-uniform sequential colormap could be used to make the figure more readable, even in black & white. Other figures could also benefit from the use of perceptually uniform colormaps, instead of the default jet. Figures 9, 10, 11, 12, 13, 14, 15. Such colormaps not only improve readability but also make figures accessible to those with different forms of color vision deficiency.
Citation: https://doi.org/10.5194/nhess-2023-210-RC3 - AC3: 'Reply on RC3', kevin cheung, 22 Apr 2024
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
450 | 163 | 45 | 658 | 41 | 36 |
- HTML: 450
- PDF: 163
- XML: 45
- Total: 658
- BibTeX: 41
- EndNote: 36
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1