Verification of avalanche bulletins by questionnaires
Abstract
The paper describes how avalanche bulletins (regional and national) are being verified by means of questionnaires in Switzerland. At the beginning of winter 1996/97 a set of four questionnaires has been introduced:
• Questionnaire A: "Personal judgement of avalanche danger"
• Questionnaire B: "Observations of avalanches without personal or material damage"
• Questionnaire C: "Observations of avalanches with people caught but without material
damage"
• Questionnaire D: "Observations of avalanches with personal and/or material damage"
About 10000 of these questionnaires have been distributed to the public, 500 have been returned during the winter 1996/97. This means twice as many returns as for previous winters. Questionnaires B,C and D can be used for the verification of the higher hazard degrees. Questionnaire A is the most frequently returned and can be used for all degrees of hazard. Therefore it is the subject of our evaluation. The questionnaires are used on a daily basis for the construction of the bulletin if they were returned in time. Therefore we cannot talk of an independent verification in a strict sense. At the end of the winter a comparison between questionnaires and bulletins has been done. The questionnaires allow to define a degree of hazard for 8 expositions and 4 height zones (lSOO-2000m, 2000-2S00m, 2500- 3000m and >3000m). Therefore we compare a set of 32 segments of every questionnaire to the bulletin. The overall estimation for Switzerland shows that about 64% of the feedback are in agreement with the bulletin, 32% differ by l degree of hazard and 4% by 2. The symmetric distribution of differences indicates that the degree of hazard is neither systematically over- nor underestimated. The distribution varies heavily between regions, height levels and expositions. For the region Davos a detailed verification using additional own field observations has been done and is compared to the verification based on questionnaires.