Most expeditions seem to rely on Google Maps – which are fairly convenient, easy to use and ofcourse is free. One of the perplexing issues is the proprietary nature of the digital maps and under scientific pretext I wanted to avoid tying down any data collected on my trip onto such platform. This has especially come to light considering the NSA PRISM scandal, although I’m not too pessimistic that I’m worried someone is spying on me.
All the data collected is stored on a (Mysql) database on the website – this includes
- The itinerary route
- Passenger fares requested and completed
- Tracking system data
The Passenger locations are plotted on the map along with the tracking data simultaenously. This will inevitably become very large and will be optimised so that it is simplified.
All registered – completed fares can opt and once completed data will be anonymised under the user’s preference. Data is only accessible to myself and will not be shared except between you and me.
Having data that is open ensures it’s accountable and any discrepencies can be found – although to be honest is not crucial on this project.
A variety of sensor data is collected from my DIY Raspberry Pi Tracking Device. It collects Latitude and Longitude GPS coordinates, temperature and elevation from a BMP085 sensor.
The sensors are polled and its speed can be changed by a standard television remote control. A moving average is performed to smooth the data and this is stored into a CSV file which can be ready by standard spreadsheet software.
If a sensor reading was taken every 10 seconds for 8 hours, that is approximatly 2900 data points. As you can imagine, that will starts to baloon over the course of one year.
GIS – Geographic Information Systems
Simple way of putting it, is plotting data onto a map and using it as a tool to correlate or find patterns between a multi-faceted topic, whether it be population density and relative clossness to coast or river systems. It’s a useful tool and will later help present all the information once completed.
Open Source GIS: Quantum GIS
Quantum GIS is a free tool for manipulating and visualising geographical data and can be run ony any platform (Windows, Mac OS X and Linux). It also offers many plugins that can be easily downloaded from within the application.
Unfortunatly the interface is fairly complex relative to Google Earth and it took a good length of time to familiarise with the basic functions. It only offers a 2D cartographic map which doesn’t have fluid panning like we are accustomed to with online maps.
Despite the issues with usability, it is a very powerful tool, to overlay many layers of data. This includes any web map service using the Open Layers plugin which can provide map from sources other than google, such as the Open Cycle Maps .
Some mapping services do allow you to generate elevation profiles from your route, but the actual presentation of elevationd data is often excluded or is very poor in respect to planning. These services rely on elevation data generated by numerous satelites. All this data is stitched together to form a grayscale image that now comprehensively covers the whole expanse of the world. Each pixel represents the elevation above sea level and the best resolution is currently available at 90 m/px. By 2014 there may even be 12m/px resolution offered in World GEM by Astriam, EADS
The GDEM – Global Digital Elevation Data can be freely downloaded from several locations – these I found most useful:
Since it’s scientific data created by NASA and ESA, it can be freely used, however, the caveat is there are no truly open services available. Google has an API aswell as Mapquest to allow you to find the elevation at any given latitude and longitude but have restrictions – unsuitable for my project.
GPS Vertical Accuracy:
The accuracy of GPS is dependant on signal strength, affected by
- interference in the ionosphere and upper atmospheric conditions
- reflected signals from tall buildings
- number of satelites in view
The resultant accuracy of GPS devices is still very good – the worst case being 7.8m – gps.gov. However the vertical accuracy is significantly worse and the worst case is upto 2-3x worse than the horizontal accuracy – USGS. This I can believe, because Eastrington, UK is approx 5m above sea level – great when the sea level rises, but my GPS was reporting 30m.
This is where GIS data can be used. I’ve conjured up a program that can read the Elevation tile data and calculate the elevation using this satelite data and uses bicubic interpolation. This is used to calibrate the sensor data. A relative altitude change is then added to from the change in elevation read by the pressure sensor on the BMP085 sensor and this should hopefully provide a more accurate reading – inspired by Digi Key – BMP085 Calibration
Using Elevation Data in GIS
As usual I have left things slightly late and I haven’t thoroughly planned my route to avoid mountainous regions.
It can even create an elevation graph using the profile tool plugin. The graph below is for the whole trip in Europe.
The very powerful thing it offers is allowing customisation of how data is visualised.
This tool I won’t be able to use on the trip most likely because it’s quite processor intensive. It will be left home so that my parents can just check the route against my elevation data.