2

Given an MGRS grid reference, for example 4QFJ123678 (100m precision), is there a good way to obtain the adjacent squares (north, south, east and west) from that origin square?

Is it better to compute the neighbors within the MGRS system, or convert to lat/long, find the coordinates 100m away in each direction, and convert back to MGRS?

1

Given the definition of the MGRS from wikipedia we know that your example 4QFJ123678 can be split up as follows:

  • 4Q is the Grid Zone (columns in a range of 1-60 and rows in the range C-X omitting I and O). As rows increase go further east, as columns increase go further North.
  • FJ is the Grid Square (columns in the range A-Z and rows range A-V, both omitting I and O). Values increase to the East and North.
  • 123 678 is the precision down to the 100 metre square in columns and rows. Values increase to the East and North.

So in the simplest case to get up, down, left, and right you just need to subtract or add from the 100 metre precision rows and columns, thus you get:

  • North 4Q FJ 123 679
  • South 4Q FJ 123 677
  • East 4Q FJ 124 678
  • West 4Q FJ 122 678

If you wanted to you could write this into a script pretty easily, but you would have to handle the cases where you cross Grid Squares and Grid Zones.

Also if you use Python might be to look at the Python mgrs library that converts to/from MGRS from latitude/longitude.

  • Thanks, I suspected that the solution might be that simple but wasn't sure about generalizing it for all cases. I've seen library code that converts the MGRS grid reference to lat/long, computes the new lat/long 100m away, and then converts back to MGRS, but that seemed a bit roundabout. – Matt Mar 11 '15 at 23:00
  • 3
    That method would make the edge cases easier (crossing latitude bands, grid squares, zone boundaries) because the lat/lon-to-MGRS function should already handle them. – mkennedy Mar 12 '15 at 16:07
0

I believe that you cannot rely on computing adjacent MGRS grid references by altering the numeric values. I'm pretty sure that method works for going North and South but West or East is a problem because not all grid squares are 100,000km wide 30U YC 051 804 has an easterly neighbour of 31U BT 804 948 for instance, instead of the computable 30U YC 052 804. It is possible that even the method of moving a search point by a set distance may fail to return the adjacent grid square if the set distance is greater than the width of an adjacent grid square. I have a requirement for a system that allocates a location with a grid square identifier and then is able to find all such locations within a radius. I was hoping to do an initial coarse grain search for locations by getting all the grid square identifiers intersecting the area. Am not sure now if mgrs is the system to use due to this problem.

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.