I am in the gsun 6383 class taught by dr lucy wong. This is our article review presentation today i will be talking about soil color mapping done with gps, smartphone tracking, as well as Music, detecting hot spots on used trails. So these devices that will be used to detect the hot spots are going to be anything wearable. Smart watches health, trackers phones or regular gps modules, handheld gps – you can call them. There will be a methodology section covering what was done to collect the data, how the data was processed, as well as how the maps were created using those data banks well go over. The results well look at a product map that was created and then well have some images showing you how the sampling was done from those databanks then well go over some conclusions. My thoughts and opinions on the article and what was done in the article so heres. The methodology they did sampling from the gps location trackers on the smart watches, the phones, the health trackers that these people were wearing and they estimated the gps accuracy and because were using gps, enabled devices. These are not mapping grade or surveying grade gps antennas. So, given that already we can assume the accuracy was between 5, you know to 10 meters, because theres, no real time, kinematics or any kind of uh real time processing happening to the location, the signal location. So they estimated that and then thats how they got the dots on the map on figure one.

They used a buffer analysis to Music in figuring out the sampling for the hot and cold spots found on the trails. They begin by estimating the gps positioning accuracy with smartphones. They need four visible satellites to get a accurate measurement. They then created a buffer analysis to distinguish movement on formal trails and informal paths. So the informal paths were just trails taken that were not created by the forest service or maintained trails, and they did that to explore spatial patterns of different recreational groups. They did a dense density analysis to locate these hot spots and they also did small scale field mapping to validate the results of the hot spot analysis in estimating the accuracy of the gps track data. They found that the smartphones were a factor in figuring out these hotspots, so they added that correction into the analysis. They used two different maps to create the uh trail maps. They use the city of helsinki roadmap and they also used a topographic database from the national land survey of finland. Of course, these two different two different maps from two different sources had varying inconsistencies, but the nls map, the national land survey of finland map, had slightly higher detail. The next step in the analysis was the proximity analysis, and i think this correlates well with the course they did this to calculate the average distance of point data of all on trail, gps tracks. They use the generate near table tool to find the shortest distance of the gps points to the trail line features within a 20 meter radius.

Of course, this is in metric units. There was a deviation on average of nine meters, so they created a 10 meter buffer along the formal trailer network, and that was sufficient for further analysis. The gps line data was then intersected with the buffered trail network to distinguish on trail from off trail movement. Youll see in the next slide what this produced. They then created a raster map using the kernel density analysis tool. This is something that weve explored in some of our arcgis trainings, and they did this for ecological and social applications in hotspot mapping and they used the density analysis tool for the gps line, features in the neighborhood of each raster cell, using a 10 by 10 meter By 10 meter raster cell size and it created a continuous surface surrounding each line based on a quadratic formula, and then they use line data to avoid bias towards spatial clustering of gps points due to participants standing still. So this is the way they negated. Smartphone location sampling, showing high traffic because the smartphone doesnt know that youre not moving to not take locations. Itll just keep taking location samples, even if youre standing still looking at your phone or taking a break drinking water, so they negated those biases. The last step was validating these calculations and analysis. They went out to the study area walked through the trails, found physical evidence of high traffic high usage, and then they also checked the off trail usage and it correlated with their findings in the smartphone gps tracking data.

So this is the map. That kind of concludes all of their analysis. They collected 55 gps tracks, 70 men and 30 women 83, with higher education and 80 in the 25 to 44 age group, were engaging in these recreational activities from the smartphone gps data. There were clear differences of how recreational movement of these groups based on their age and gender were distributed on formal trails and informal paths. Runners mainly followed formal trails, with only 21 percent of the gps tracks being off trail, while forty first forty, six percent of mountain bikers tracks were located outside the formal network. Basically, what theyre saying is that most people that were running stayed on the trails, those that had mountain bikes seem to veer off trail and venture. So this was a important study done to highlight some of the uses of kernel density mapping as well as buffer zones and the analysis process used. There was only an eight meter, deviance of result, accuracy between the smartphone, the health trackers and the watches used to collect the data so heres, some of the in conclusion here, some of the pictures of physical evidence showing how the smartphone data correlates to physical on the Ground observations in the pictures. You can see trails from mountain bikes. You can see shrub growth weaning, whereas its not as dense, because people are moving through it, and so one of the things they did figure out in the end was that theyre not sure of the motives of why people were moving off the trails, but they did Show that there was a correlation between mountain bikers age groups and certain recreational activities veering off the trail and then using our geospatial analysis processes to show that that is the end of my presentation.