0

I have created a point --> Figure 1

enter image description here

I have then selected the Buffer tool and created a buffer around my chosen point, with a degree interval of 0.012. I have selected the End style cap as square and run in background. This has produced a buffer around my point

Figure 2

enter image description here

Because I am using WGS84 CRS, I cannot use meters as the chosen map unit, instead I am using degrees and I have to guess a degree to match close to what I want. Does anyone have a solution to this specific problem?

My next problem is that after trial and error is that once created a buffer around the right dimensions they are constantly providing contradictory data.

Figure 3

enter image description here

From the right side it records a distance of 705m until the center, giving a diameter of around 1400m

Figure 4

enter image description here

From the left side it produces a distance to the center point of 1343m until the center and a diameter of around 2700m.

Figure 5

enter image description here

indicates that from corner to corner the distance is almost exactly 3000m.

Can someone please explain to me how buffers work. because I am using the Buffer tool the way people are telling me too, yet the results from it seem so inaccurate.

RyanKDalton
  • 23,068
  • 17
  • 110
  • 178
Louis Tate
  • 807
  • 1
  • 8
  • 22
  • 4
    Save your point-layer using a fitting UTM-projection, then create the buffer, then - if necessary - save back to WGS84. Problem solved. – Erik Nov 26 '18 at 15:33
  • what do you mean by 'fitting UTM-projection' what is classified by this? – Louis Tate Nov 26 '18 at 15:35
  • if you mean its the CRS I want my point data to be displayed in, then I have already done that, and the error still occurs – Louis Tate Nov 26 '18 at 15:35
  • Depending on where your data is situated, you have to choose the correct UTM-strip. Do all buffering using this CRS. Then measure again. – Erik Nov 26 '18 at 15:38
  • Ok, I have originally used ED50 31N CRS, then changed the map, however when I make the buffer back in the ED50 31N, it does work as you said it would. Now all I need to do is save the buffer back in WGS84, then change my map projection back to WGS84? – Louis Tate Nov 26 '18 at 15:39
  • What @Erik is saying is to use a webpage like this to see which UTM Zone would work for your data. Reproject your data into that UTM Zone so that you can buffer accurately in metres. Then reproject back into WGS84 if that's how you want your project displayed.

    Also the reason you are getting different distance results north vs east is because your square is not actually a square, it is a rectangle that looks like a square because it is being projected in WGS84.

     – TeddyTedTed Nov 26 '18 at 15:41
  • So the imbalance in distance is not so much an issue as it is being accurately displayed within the CRS that I want. Also thank you both of you for clearing up this issue – Louis Tate Nov 26 '18 at 15:43
  • 5
    What seems like a square in WGS84 actually is not a square in real life. – Erik Nov 26 '18 at 16:03
  • 1
    Note that you cannot just change the map projection--that only affects the display. You need to use the reproject tool to actually reproject your points. – Jon Nov 26 '18 at 16:08
  • where do I find the reproject tool? – Louis Tate Nov 26 '18 at 16:12
  • What GIS software are you using? – PolyGeo Nov 26 '18 at 18:31
  • Possible duplicate of Create buffer in meters? – RyanKDalton Nov 26 '18 at 19:43
  • using QGIS version 3.2 – Louis Tate Nov 27 '18 at 08:56

1 Answers1

1

Because WGS84 (Lat/Long) is a Geographic Coordinate Reference System (CRS) and the distance of a "degree" gets smaller as you get closer to the Poles, you will never get an accurate buffer based on a specific "measured" distance (like meters or feet).

In QGIS, you'll want to convert your layer (as shown here) to a "Projected" CRS before you buffer the features (one which preferably is "centered" as closely as possible on your geographic extent and uses the unit of measure you are interested in). When you run the buffer routine, you should end up with a perfect circle of the same distance, when viewed in that CRS.

However, if you load that "perfect circle" into a map using a Geographic CRS, the circle will look flattened out, like an oval, as shown here. This distortion is inherent in GIS and mapping, when we try to depict a sphere (earth) on a 2D-plane. As an example, try unpeeling an orange and trying to flatten it out on the table... it is impossible for the "peel" to remain un-distorted in shape or distance.

RyanKDalton
  • 23,068
  • 17
  • 110
  • 178