161 How to elaborate precise visitor numbers? Reto Rupf, Michael Wernli, Ruedi Haller Abstract — Visitor numbers and visitor distribution are required information for various management tasks of recreational areas. Automatic data collection is a low-price opportunity to obtain data. The main problem of auto- matic methods is data precision. In the Swiss National Park, automatic visitor counting with acoustic slab sensors started in 2005. As precision did not appear to be satisfactory, the counting problems were investigated. Sensor installation strictly followed the instruction manuals given and fne tuning during the calibration period led to a devi- ance of 5%. Precise counting data resulted in the sum of persons counted whereas direction separated data was not as precise. Recommendations for counting site selection, installation and calibration counting are given. Index Terms — Acoustic slab sensor, calibration, visitor census, visitor monitoring, Swiss National Park —————————— u —————————— 1 IntroductIon V isitor management in nature parks requires a good data of visitor num- bers. Effects on regional economy, carrying capacity, necessary infrastructure, disturbance of wildlife etc. can only be esti- mated from reliable data e.g. [1] (based on survey and estimated visitor numbers form 1993 [2], Küpfer [3] estimated the economic effect of the National Park on the region of about 17.4 Million Swiss Francs). The qual- ity of visitor experience infuences a visitor’s attitude towards nature and the environment. Positive experiences promote understanding of the need for nature conservation [4]. In the Swiss National Park, the frst visitor counting and survey took place in the nineties [2], [5]. In 2006 and 2007 surveys of visitor structure and requirements were carried out [6], and the Swiss population’s perception of the National Park was ascertained [7]. In 2005 automatic visitor counting started in the Swiss National Park. Acoustic slab sensors recording hikers were chosen because they are imperceptible and use little energy [8]. However, the preci- sion of the visitor counting was not satisfac- tory. On-site precision did not correspond to counting under lab conditions [9]. Reasons for this lack of accuracy are discussed in Ross [10]. Therefore, the aim of visitor count- ing during the hiking season 2007 was to en- hance the counting precision. 2 Method During the 2007 season, an improvement in counting certainty was sought to reduce the difference between automatic and manual counting to less than 10%. Three calibration loops were made. Each loop lasted one month and contained 2-8 hours of manual counting per sensor. Loop results were analyzed and followed by system adjustments if required. Exact manuals for the installation (based on the manual of the producer Eco-Counter [11]) and calibration counting were written. Eight acoustic slab sensors from the eco ———————————————— R. Rupf, Zurich University of Applied Sciences, Insti- tute of Natural Resource Sciences, P. O. Box, 8820 Waedenswil, Switzerland. E-mail: reto.rupf@zhaw.ch M. Wernli, Zurich University of Applied Sciences, Institute of Natural Resource Sciences, P. O. Box, 8820 Waedenswil, Switzerland. E-mail: michael. wernli@zhaw.ch R. Haller, Swiss National Park, Chasa Planta Wilden- berg, 7503 Zernez, Switzerland. E-mail: rhaller@na- tionalpark.ch