Avalanche size estimation and avalanche outline determination by experts: reliability and implications for practice
Abstract. Consistent estimates of avalanche size are crucial for communicating among avalanche practitioners, but also between avalanche forecasters and the public, as for instance in public avalanche forecasts. Moreover, applications such as risk management and numerical avalanche simulations rely on accurately mapped outlines of past avalanche events. Since there is no widely applicable and objective way to measure avalanche size nor to determine the outlines of an avalanche, humans estimate these. Therefore, knowing about the reliability of avalanche size estimates and avalanche outlines is essential as errors will impact applications relying on this kind of data. Conducting three user studies, we investigate the reliability in avalanche size estimates and avalanche outlines either mapped from oblique photographs or from remotely-sensed imagery. We compare size estimates for 10 avalanches made by 170 avalanche professionals working in Europe or North America, the mappings of six avalanches from oblique photographs from 10 participants, and the mappings of avalanches visible on 2.9 km2 of remotely-sensed imagery in four different spatial resolutions from five participants. We observed an average agreement in the avalanche size estimated by the majority of respondents of 66 %, while agreement with the avalanche size considered «correct» was 74 %. Moreover, European avalanche practitioners rated avalanches significantly larger for eight out of 10 avalanches, compared to North Americans. For the outlines mapped from oblique photographs, we noted a mean overlapping proportion of 52 % for any two avalanche mappings and 60 % compared to a reference mapping. The outlines mapped from remotely-sensed imagery had a mean overlapping proportion of 46 % (image resolution 2 m) to 68 % (25 cm) between any two mappings, and 64 % (2 m) to 80 % (25 cm) when compared to the reference. Assuming that participants are equally competent in the estimation of avalanche size or the determination of avalanche outlines, we calculated a score describing the factor required to obtain the proportion of agreements between any two size estimates or overlap in avalanche areas. This factor was 0.72 for avalanche size estimates in our data set. It can be regarded as the certainty related to a size estimate by an individual, and thus provides an indication of the reliability of a label. The presented findings demonstrate that the reliability of size estimates and of mapped avalanche outlines is limited. As these data are often used as reference data or even ground truth to validate further applications, the discovered limitations and uncertainties may influence results and should be to be taken into account.
Elisabeth D. Hafner et al.
Status: open (until 22 Jun 2023)
- RC1: 'Comment on egusphere-2023-586', Ivan Moner, 29 May 2023 reply
- RC2: 'Comment on egusphere-2023-586', Brian Lazar, 07 Jun 2023 reply
Elisabeth D. Hafner et al.
Elisabeth D. Hafner et al.
Viewed (geographical distribution)
The article deals with a subject of great interest in my opinion, both for the public and for snow and avalanche practitioners. It does so in a very structured way, providing new data that are treated with appropriate and rigorous statistical methods.
The methods used are explained at length, and these lead to clear results, from which important and well-justified conclusions are derived. The methods and techniques used would allow the study to be reproduced if this were of interest.
Both the title and the abstract perfectly reflect the content of the article and are easy to understand. Formulae, symbols and abbreviations are correctly explained. The same applies to the references, which are well documented.
The article is well-structured and clear. As it deals with three different experiments, it is quite long, but this is justified. Most of the figures are easy to interpret, except for Figures 2 and 3, which are practically illegible due to the small size of the fonts and photos. We suggest increasing the size of the font and photos, at the cost of eliminating the "medium size" graphics, which are also not thoroughly worked into the text.
Overall, the article is written in clear, well-structured and simple English.
In our opinion the article can be published as is, but if you wish you can take into account the following comments:
- Line 114: We find it more appropriate to speak of "public forecasting" than "regional forecasting".
- Line 121: We are a bit surprised to find 10 responses in Spain and not having known about this study before. Knowing which organizations in each country were surveyed would give information on the quality of the survey.
- Line 124: In Study 2 the sample of 9 people is obviously small. The uniform background of the respondents further weakens the results. The same happens in Study 3. The need to improve this with a larger sample could be pointed out in the conclusions.
- Line 138: Delete the first comma
- Line 190: Figure 1a, which does not exist, is cited.
- Figure 5: We are very surprised by the result shown in this figure, in which the destructive potential is the least valued of the factors. The possible causes of this anomaly are briefly discussed in the text.
- Figure A1 does not seem relevant to us
In any case, congratulations on the work done, it is excellent to us!