Quicker mapping of Street Trees

OpenStreetMap with Observado tree observations KML

I have always been cautious about mapping trees on OSM - not because I don't want to, but because it is time-consuming and difficult to achieve consistency in mapping over reasonable areas. During 2008-9 (before I did OSM) I mapped about 100 Oak, Quercus robur, trees at Attenborough Nature Reserve. This took 5 separate visits lasting on average about an hour and a half. Of course I was doing other things as well, but each tree took about 5 minutes (I was also measuring the circumference (more properly Breast Height Girth). I still find this data set very useful (see below), but I did learn how time consuming the process can be.

Location of Oak trees at Attenborough Nature Reserve
(base map © Ordnance Survey OpenData).

One of the problems has been collecting additional data: particularly the type of tree. For a while I used a code for Garmin waypoints (a location letter, the first two letters of the genus , and the first three letters of the species and two numeric characters, e.g. TQurob01), but found it took too long to enter.

Since last September I have been using an android app called ObsMapp, created for a couple of Dutch wildlife organisations, SOVON and Waarneming. This app provides a very good basis for quickly recording almost any type of animal ,plant or fungus one might come across. Normally it uses the phone's GPS, but it can use OpenStreetMap data downloaded to the phone. One restriction is that one has to use pre-built species lists: and most of the extensive lists I use are Dutch rather than British so that from time to time I can't use it to record something. Oddly the app is closed source, but the data generated is pretty open. It's a shame that the app is not open as there are a number of enhancements which would make it more useful for the UK (e.g., using British National Grid: the base app can use the Dutch Grid, and Nick Whitelegg has an android module based on proj4 which does the grid transformation).

Data need to regularly uploaded to one of the host websites (observado.org outwith NL & BE, waarneming.nl for the Netherlands, waarneming.be for Belgium). Once uploaded they become available to view (unless prior restrictions have been added regarding location or publication date) by anyone within around 15 minutes.

Last Sunday I decided to walk into the middle of Cambridge and en route decided to use ObsMapp to record street trees for a few minutes (also limited by how fast my hands got cold!). I collected observations to species level on 30+ trees in about 15 minutes, with the excess of data collection over walking being about 10 minutes, or about 3 trees a minute. This wasn't bad at all AND collecting the species names wasno trouble at all. (It is possible to create custom check lists online at the Observado website, so this process could be speeded up: I have about 30000 biota in the checklist on the phone).

Transferring the data I collected to OSM was slightly less straightforward. Firstly, it's only possible to download one's own data from observado. Secondly, only CSV and KML formats are supported. KML is nice for visualisation, but one loses the detail of the record, so I went with CSV. This also presents a problem because the standard CSV delivered uses a Dutch locale, so I had to change the locale in the software I used to read the file (in practice quite easy with OpenOffice). I then read this into QGIS and saved as a shapefile. Finally I used pnorman's version of ogr2osm to create an OSM file which I loaded into JOSM and hand edited tag names. I also changed all occurrences of Japanese Cherry (Prunus serratula) to Flowering Cherry. Most of these street cherry trees were grafted on a root stock of Wild Cherry Prunus avium, but are likely to be different cultivars flowering at different times. Although most ornamental cherries have some serratula genes, their origins are now so obscure that they are not given specific names, but referred to using the cultivar name after the genus as in Prunus 'Kanzan'.

One problem that does arise is that in this area, the pavements (sidewalks) have been mapped as separate footways. Personally, I dislike this kind of mapping: it messes up pedestrian routing and adds more detail than is necessary. Presumably this was done because these roads have grass verges. The verges are very narrow (no more than a metre), but because the street trees are planted in the verges, my GPS points are sometimes shown on the wrong side of the path. I don't actually believe that these footways are mapped accurately enough to move the trees to show the actual topological arrangement.

Although only mapped about half the trees on the couple of streets I chose as a testbed, I was able to map consistently. It will be relatively straightforward to collect the remaining information. Although the process of post-collection data manipulation is a bit tedious, I would still recommend ObsMapp for collecting tree data for OSM, it's a lot faster and less error prone than anything else I've tried.