In this study, we tried to analyze user data of a mountain hiking app, but it was necessary to limit the analysis targets because the data collected through the app had a wide range, a large amount, and fast accumulating speed. In a social atmosphere in which exercise is common (Lee, 2018), the Generation MZ is newly expanding and linking the existing range of exercise for complex reasons such as the tendency to feel a sense of accomplishment through exercise and the growth of the related device and platform market (Kim et al., 2020). Generation MZ or MZers is a term that encompasses Generation M and Z: the generation born in the early 1980s to mid-1990s are classified as millennials (Generation M), and the one born in the mid-1990s to mid-2000s as Generation Z (Univtomorrow20slab, 2020). To examine the forest usage patterns of MZers who actively participated in exercise, such as the recent increase in the proportion of use of exercise platforms, the age range was set for those in their 20s and 30s, the age group corresponding to being the MZ generation. In addition, referring to the findings of Park and Kim (2001) that mountain forest type had a high usage rate of middle-aged people and suburban forest type had a high proportion of young people in their twenties, the range of the place was set to 13 mountains with high frequency of visits in Seoul metropolitan area.

Beagle, the developer of the workout app “Tranggle,” has collected big data specialized for outdoor activities and exercise, and has provided various services based on artificial intelligence (Beagle, 2021). Since 2011, Beagle has built a tracking big data center that includes a vast amount of information such as 23 million outdoor exercise courses and tracking course data, provided by 1.23 million Tranggle GPS members, and more than 20 additional items of information, which have been used in the field of big data business (Chosun Biz, 2019). The data used in this study were provided on those in their 20s and 30s who visited the top 13 most frequently visited mountains in the Seoul metropolitan area, including Mt. A-cha, Gwan-ak, Gwang-gyo, In-wang, Bul-am, Su-rak, Do-bong, Sa-pae, Dae-mo, Sam-sung, Cheong-gye, Buk-han, and Namhan-sansung, through approval procedures after consulting with the relevant company. The period of data was from January 2019 to early December 2020, and the total number of data was 337,504, consisting of 142,806 in 2019 and 194,698 in 2020. Statistics on members in 2018–2020 and analysis data of hiking activities of those in their 20s and 30s based on big data of Tranggle GPS, provided by Beagle, were also referenced. For the forest area by “Gu”, an administrative district in Seoul, statistics on the forest area of Seoul (by forest type) provided by the Seoul Open Data Plaza were used.

The data provided had a total of 22 variables, such as classification, experience point, actual exercise distance, average speed, maximum speed, total time, travel time, maximum altitude, registration date and time, and region 2. We used these variables to create derived variables. From the classification, each group such as Hiking, and Walking was extracted. To increase the fun of mountain hikes, Tranggle has applied a ranking according to users’ experience point and given a level to them based on the ranking (Tranggle, 2021); experience points are given only to the effective exercise record (actual distance traveled) for hiking tracks, and the unit is mileage (M) (Tranggle, 2021). By referring this, “Level” was created by dividing experience points into sections based on the Tranggle level (rank) standard table. As registration date and time were divided into year, month, day, hour, and day of the week, “Year”, “Month”, “Day”, “Hour”, and “Weekday” were created. For month, “Winter season” and “Summer season” were created by referring to the criteria of winter season (November-March) and summer season (April-October) classified by Korea National Park Service (KNPS). “Hour” was divided into working hours (09–18 hours), night (19–23 hours), early morning (24–8 hours), and all (24 hours) according to the definition of daily routine (regular tasks done every day (National Institute of the Korean Language, 2021)). In addition, “Rest time” was created by subtracting travel time from the total time; and “Region 2 (city/province)” and “Region 3 (city/province)” were created by dividing region 2 according to the unit of administrative district. Based on this, data for the “Metropolitan area” group were created by combining data from Seoul, Gyeonggi-do, and Incheon. In addition, a “Mountain name” variable was created according to each file name provided. Afterwards, processing into necessary data was carried out by checking and removing missing values and outliers.

The 13 mountains with the highest frequency of visits in the metropolitan area, which were derived from users in their 20s and 30s using Tranggle app, were mostly located in the northern and southern outskirts of Seoul, and were suburban forests that form the framework of the city (Fig. 2).

In 2020, the number of visits to mountains in the metropolitan area by those in their 20s and 30s using Tranggle app increased 1.36 times compared to 2019. This result is in line with the reports, such as the statistics of visitors to national parks showing preference for suburban forest hikes over long-distance travel due to the COVID-19 pandemic (Korea National Park, 2020); and the findings of travel patterns and behavior of domestic tourists using SKT’s T-map traffic data and KT’s big data, indicating that demand for short-distance tourism around metropolitan areas and large cities increased rather than long-distance travel (Korea Tourism Organization, 2020; Ministry of Culture, Sports and Tourism, 2020).

To examine the factors that increased mountain hikes in the metropolitan area, we analyzed the top 3 regions (si (city), gun (county) and gu (district)) with high frequency of usage by mountain (Table 1). For the mountains including Buk-han, Bul-am, Su-rak, and Cheong-gye, the frequency of usage by region was evenly distributed in the past, or spread evenly in 2020; and for the other nine mountains, except for the three mountains, the frequency of usage in the regions, which showed the highest frequency of usage in 2019, got higher. This was related to the regional distribution of hiking trails by mountain. For mountains where the regions used by hikers were evenly distributed, hiking trails were evenly created in nearby areas, and for mountains with high frequency of use only in specific regions, hiking trails were built around specific areas only. In addition, according to the analysis results of hiking activities by users in their 20s to 30s based on big data acquired from Beagle’s Tranggle (Beagle, 2020), visitors to Mt. Buk-han were evenly distributed over many peaks of the mountain, and those to Mt. Gwan-ak were concentrated on the Sadang Station-Yeonjudae course, and those to Do-bong on the three main peaks of the mountain. Those in their 20s and 30s who used Tranggle used mountain trails actively, showing that their preferred hiking sections can be divided by mountain.

To look into the pattern of mountain hikes in the metropolitan area in a compressed manner, the regional range was reduced to Seoul, where users accounted for more than 50% of the population. In addition, the forest area and research data by gu in Seoul were classified by dividing into All (all classifications), Cycling, Walking, and Hiking, and the number of usages for each classification in Seoul in 2019 and 2020 were counted (Table 2). The results of Spearman correlation analysis conducted with the counted data showed high correlations such as in 2019, All (0.8, p < .001), Walking (0.69, p < .001), and Hiking (0.81, p < .001), and in 2020, All (0.76, p < .001), Walking (0.57, p < . 01), and Hiking (0.76, p < . 001) (Table 2). It can be considered that the larger the forest area of a gu, the more forest usages (All), Walking, and Hiking are done in the district. This is in line with the area of urban forests (Park et al., 2008), which was mentioned as a factor to increase the frequency of forest visits in previous studies. However, in 2020, the Spearman correlation coefficient declined compared to 2019, despite the fact that the larger the forest area in a gu, the higher the frequency of usage in All, Hiking and Walking. In Seoul, there are eight Seoul Dulle-gil (walking trail) courses that connect the outskirts of Seoul, which were among the top 13 hiking trails with the highest frequency of visits by hikers in their 20s and 30s in the metropolitan area. In addition, the top 5 mountains in the metropolitan area with the highest number of visits compared to 2019 were A-cha, Gwan-ak, Gwang-gyo, In-wang, and Bul-am in descending order; A-cha, Bul-am, and Gwang-gyo are easily accessible on foot, and Buk-han and Gwan-ak are also highly accessible as those mountain trails are well connected within the city (Beagle, 2020). Based on these data, it can be considered that hikers in their 20s and 30s using Tranggle in 2020 still had a high number of visits to a “gu” (district) with a large forest area, but they more frequently used forest recreation areas with convenient transportation or easy access to the mountain within the city center. In addition, after the COVID-19 outbreak, it has been observed that people’s trips to nearby areas increased, making them a single tourist zone, even if the administrative district (metropolitan city, or province) to which their residential area belonged was different from that of a desired destination (Korea Tourism Organization, 2021a). Considering the above results, the comprehensive analysis of hiking regions can be summarized as follows: during the COVID-19 pandemic, the hiking tendency of those in their 20s and 30s using Tranggle was still high in terms of demand for a large forest area, but its influence was lower than in 2019; and the accessibility to and convenience of using the mountains within their living area were regarded as important. Furthermore, it was found that three of the top 5 mountains with increased frequency of visits in 2020, including A-cha, Gwan-ak, and In-wang, are located in downtown Seoul, or have convenient public transportation connecting them to the city center. It is thought that not only hiking but also commercial areas around such mountains were considered and combined with tourism.

The total number of accumulated data on mountain hikes by users in their 20s and 30s in the metropolitan area by year was a total of 337,504, such as 142,806 in 2019 and 194,698 in 2020, showing an increase of 1.36 times in 2020 compared to in 2019. The hiking pattern was analyzed by dividing the temporal factors by month, day of the week, and time of day based on the year.

To classify each month according to the time of sunset, it was divided into Winter season (November-March), Summer season (April-October), and All (12 months), and the high or low usage for each classified period was determined as the mean. Comparing the two years, in 2020, the mean was lowered due to data collected only until the beginning of December, so the number of usages in February appeared to be higher than the mean in the Winter season (Table 3). Spring and autumn showed high frequency of usage, which included months with higher number of usages than the mean for Winter season, Summer season, or All.

Mountain hikes are greatly influenced by the weather due to the nature of outdoor activities. In particular, in 2020, abnormal weather caused by the climate crisis occurred frequently along with COVID-19 (Korea Meteorological Administration, 2021). It is necessary to look at the monthly hikes in 2020 in consideration of the weather characteristics. The winter of 2020 (December 2019-February 2020) was warm nationwide, especially in January, with the warmest temperatures since 1973. In addition, since 1973, the average temperature in spring has severely fluctuated in March (top 2), April (top 44), and May (top 14), but the number of days with yellow dust in spring was 1.0 days, which was less than the normal year (5.4 days). (Korea Meteorological Administration, 2020a). Weather characteristics such as warm temperature and pleasant atmosphere seem to be one of the factors that increased mountain hikes in winter and spring. After that, the number of usages decreased due to frequent rain that fell once every 3 days in May and the heatwave in early June; but the number of usages in June was maintained in July rather than decreasing, due to the lower temperature in July than in June. After mid-August, their mountain hikes seems to have decreased significantly due to heatwaves and tropical nights (Korea Meteorological Administration, 2020a; Korea Meteorological Administration, 2020b). The difference between 2019 and 2020 in the time of decrease in mountain hikes in summer and that of increase in hikes in autumn appears to have resulted from: the temperature fluctuations in summer, 2020, which were different from common years with heatwaves from the end of July to the mid-August; and delayed mountain hikes in autumn due to the longer rainy season that lasted about two months (Korea Meteorological Administration, 2020b).

The days of the week were classified by dividing into Weekdays (Monday-Friday), Weekends (Saturday-Sunday), and All (Monday-Sunday) and the means were calculated. Comparing the two years, in both years the mountain usage was highest on the weekends and higher on Saturdays than Sundays. In addition, among Weekdays, the days regarding which the number of usages was above the mean changed from Tuesday-Thursday to Wednesday-Friday (Table 4). Fig. 3b shows more clearly the difference between the two years in the number of usages on weekdays. In 2019, the slope of the graph was large due to the highest usage on Wednesday, but in 2020, the slope was modest due to the small difference in usage between weekdays. To interpret this result, it is necessary to refer to the decrease in the trip/travel of those in their 20s and 30s after the outbreak of COVID-19 based on mobile big data (Statistics Korea, 2020). It is considered that the decrease in the difference in mountain hikes between weekdays in the age group in 2020 was attributed to their increase in mountain hikes in nearby areas due to the reinforcement of quarantine rules, such as social distancing and refraining from private gatherings, despite their spare time increasing due to the spread of telecommuting and remote learning caused by the COVID-19 pandemic.

Usage by time of day was divided into Work time (9–18 o’clock), Night (19–23 o’clock), Early morning (0–8 o’clock), and All (24 hours). Comparing the two years, period of time with above the mean of Work time and All were 13–17 o’clock and 11–18 o’clock, respectively, which did not change in both years. However, the time of day with above the mean of the Early morning and Night advanced from 7 o’clock to 6 o’clock and from 21 o’clock to 19 o’clock, respectively (Table 5). And despite the increase in the number of data in 2020 by 1.36 times, the number of usages at 0–4 o’clock and 20–24 o’clock was similar or lower than that in 2019, indicating a decrease in usage. This change was also confirmed in the pattern of shopping time paid by credit cards after the COVID-19 outbreak. Shinhan Card’s Big Data Center reported that the percentage of 10–13 o’clock among users’ shopping time in large-sized supermarkets on weekends increased by 3% compared to 2019, showing that they seemed to avoid using them during times of big crowds. Pointing out that the proportion of evening hours among the hours of use of convenience stores increased, and related words mentioned on social media such as “home”, “beer”, and “supper” were ranked at the top, it also reported that the time of returning home and having meals advanced due to the atmosphere of refraining from private gatherings as much as possible (Shinhan Card, 2021). It appeared that such considerations and influences affected mountain usage for hikes, so their use early in the morning when there were few people on the mountains increased, and the usage at night decreased due to earlier time of returning home.

Looking at the top 3 activities of users in their 20s and 30s, the rate of use for Hiking and Walking was still high, and Cycling increased (Table 6).

Of all the types of activities, comparing the mountains with the highest increase in number of hikes compared to 2019, including A-cha, Gwan-ak, Gwang-gyo, In-wang, and Bul-am, and the hiking trails most mentioned in the social big data analysis by Seoul Tourism Organization (2020), In-wang and A-cha were both highly counted. In addition, after the COVID-19 outbreak, the top 3 mountains with high frequency of visits for the purpose of Hiking (2019: Buk-han, Gwan-ak, Do-bong, 2020: Gwan-ak, Buk-han, Do-bong), Walking (2019: Buk-han, A-cha, Bul-am, 2020: Buk-han, A-cha, Namhan-sansung), and Cycling (2019, 2020: A-cha, Sam-sung, Buk-han) did not change significantly (Table 7). However, the fact that there was a difference in the top 3 mountains with high frequency of visits between activity types shows that the preferred mountains may differ for each activity type. Mt. Buk-han is considered to be a mountain that covers a variety of activities as it was counted as the highest frequency of visits in all of Hiking, Walking, and Cycling. The reason why Mt. A-cha was counted as one of the top 3 mountains in Walking and Cycling seems to be that such activities can be performed smoothly due to its low altitude (287m). Such difference in the preference for mountains by activity type was reconfirmed in the following findings: when analyzing the words associated with Seoul’s mountains and Dulle-gil courses on social media, those for Mt. Buk-han were related to mountain hikes focused on active hiking, such as each peak and hiking gear, while those for Mt Gwan-ak were related to lighter hikes, such as daily life, scenery, and rest (Seoul Metropolitan Government, 2018).

When converting the yearly experience points of users in their 20s and 30s who visited the mountains in the metropolitan area according to Tranggle level standard table, about 90% of them were distributed in the lower six Levels in both years (Table 8). Compared to 2019, the experience points of Level 0 and Level 1 decreased in 2020, and this appeared to result from even more new subscribers (Level 0) in 2019, although there were significant increases in them in August-September 2019 and February 2020. In addition, it is regarded that the proportion of Level 2-Others has increased as existing subscribers’ steady mountain hikes (Fig. 4) increased their experience points.

Then, the mean of the number of usages of each Level was calculated for each year, and the mountains in which the number of usages for each Level was above the mean were classified (Table 8). The mountains where the number of usages for each Level was above the mean after the COVID-19 outbreak showed no significant change except for Mt. Bul-am. However, it was necessary to pay attention to the distribution of the Levels by mountain. Mt. Buk-han and Gwan-ak had relatively high altitude among the 13 mountains, but all Levels were evenly distributed in the two mountains. Mt. Do-bong, Su-rak, Cheong-gye and Gwang-gyo were mountains higher than the average altitude (538 m), and were frequently visited by Level 4-Others, and followed by Level 0 and Level 1. Mt. Bul-am and Sam-sung are slightly lower than the average altitude of the 13 mountains, and appear to have mid-altitude between high and low mountains. Mt. In-wang, A-cha, and Dae-mo were also slightly lower than the average altitude, and were frequently visited by Level 0-Level 3 (Table 8). This result shows that the difficulties in mountain hikes varies according to the altitude, leading to the characteristics of each mountain. Hikers in their 20s and 30s using Tranggle are considered to visit mountains selectively according to these mountain-specific characteristics and their own physical abilities.

Finally, since the median is appropriate as a representative value of the data with singular values, the characteristic of each mountain hike in both years was compared based on the median (Kim, 2009). The experience point was 834 m in 2019, and 916 m in 2020; the actual exercise distance was 8,698 m, and 9,180 m, respectively; the average speed was 3 Km/h, and 3 Km/h; the top speed was 7 Km/h, and 7 Km/h; the total time was 247 minutes, and 257 minutes; the travel time was 208 minutes, and 213 minutes; the break time was 28 minutes, and 33 minutes; and the highest altitude was 610 m, and 628 m. Excluding speed, the medians of all characteristics increased compared to in 2019 (Table 9). According to the analysis of mountain hiking activities in those in their 20s and 30s based on Beagle’s Tranggle Big Data, the average hiking time was 4–5 hours, the average hiking distance was 6.62 km (3–4 km one way), and the average number of hikes per month was 0.6. It was reported that the characteristics of mountain hikes of young people in their 20s and 30s were different from those of middle-aged people (Beagle, 2020). Although the number of data increased by 1.36 times, the representative values of the exercise characteristics did not increase significantly. The rest of the items were similar to those in 2019, except that the experience point increased by 1.1 times and the break time increased by 1.18 times (Table 9).

Jang et al. (2019) classified 0–1 m/s as stopping walking and staying, 1.0 to 1.5 m/s as normal walking, and 1.5 m/s or more as fast walking by considering the topographical characteristics of mountain areas. For those in their 20s and 30s using Tranggle, as the average and maximum speed were 0.83 m/s and 1.94 m/s, respectively, it is considered that they alternated between normal walking and fast walking.

Yoo et al. (2013) found that the rate of hikers spending longer than 3 hours was 73.2% for Mt. Buk-han, and 64.4% for Mt. Gwan-ak. Yoo et al. (2004) reported that the amount of time spent in an urban park-based urban forest in the city center was about 1–2 hours, and that in an urban forest based on natural ecology outside the city was about 2–4 hours. According to Lee et al. (2009), 48.1% of visitors to Mt. Buk-han on weekdays and 64.8% of visitors on weekends stayed for more than 2 hours. It is considered that the hiking time of those in the 20s and 30s using Tranggle was similar to or slightly more than that of all ages.