I'm using the Smithsonian Global Volcanism Database and have a .csv lat lon points from it. It seems they are using decimal precision on their lat lon to give a rough idea of the size of the area that a volcanic center covers. For example, the enormous Phlegrean Fields are given as 37.1, 12.7 while a single volcano like Vesuvius is given as 40.821, 14.426. At least I think that's what they're trying to do (and yeah, I realize that would be a bad way to do things).

Anyway, what I want to do is create polygons that cover the entire area specified by the imprecise locations for each volcano. I mostly use QGIS and related open source tools and I could probably script this in QGIS python pretty easily, is there an existing tool I can use to do this instead?

EDIT: Note that often the latitude precision does not match the longitude precision. For example, Kurab volcano is given as 11.88, 41.208 .


In your csv, create a new 'size' field that counts the number of decimals in the lat/long fields. Use that as a sort of key to buffer the points at variable distances (and hence, create your polygons of different sizes).

EDIT for less vagueness:

  1. In Excel or similar, extract the number of decimal places into a new column (let's call it 'size' since you are inferring that the number of decimals is the inverse of the size of the volcano/area).

    Volcano | Lat | Long | Size
    Vesuvius | 40.821 | 14.426 | 3
    Phlegrean |37.1 | 12.7 | 1

You now have a relative size indicator you can use to size your polygons.

  1. I would then alter the values in the 'size' column to be the size you want to buffer the points (e.g. 1=10km, 2=5km, 3=1km)

  2. Load your points into QGIS/ARC/etc.

  3. Open the 'Buffer' tool from the geoprocessing menu.

  4. Select your layer and select the 'Buffer distance field' option. Set this to your 'size' column

When you run this you'll have the variable sizing of polygons you wanted. You noted that the lat and long do not always share the same precision. Pick one and stick with it. I do think it is a pretty big assumption that the decimal places correlate to the size to begin with, but I am assuming you have quality checked that. I'm just here to provide a solution to your problem.

  • Your answer is quite vague. Can you explain the technique you are suggesting in more detail, please? – foobarbecue Jul 5 '16 at 20:49
  • Ok, I see what you're suggesting now. Upvote for effort. Your solution doesn't work for me because I'm looking for the precise area which is specified (rectangles) whereas you're talking about drawing "back of the envelope" circles. I'll just script it in python. Surprising there isn't something out there already considering coordinate pairs actually describe rectangles of possibility when you consider significant digits... – foobarbecue Jul 5 '16 at 21:46
  • I think maybe I misunderstood the data you have. Based on your description you have points (presumably in the centre of the volcano?). If that is the case, would circles not be a better representation than rectangles? – rbhamjee Jul 5 '16 at 21:50
  • To understand my question, ignore the specific application -- that's just an example. Consider a coordinate pair like 10.1, 20.22 . You might think of that as a point, but it's actually a rectangle with one corner near 10.05, 20.215 and the other near 10.15, 20.225. My question is "what's the easiest way to make a shapefile of those rectangles." If you're curious, some of coordinates represent a single volcano, but others represent a "volcanic center" / "volcanic complex" which can have many cones and craters. Yellowstone, for example, has about 6 major eruption craters. – foobarbecue Jul 5 '16 at 22:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.