- Provides an objective method of measuring spatial distribution where a visual method would be rather subjective.
- Need to define what is meant by a settlement (individual houses or a cluster of buildings greater than a prescribed number).
- Need to decide where you measure from - the edge of the settlement or the centre.
- Need to decide whether to measure straight line or road distance.
- The Rn index can obliterate important sub patterns within the overall area.
- Rn index will not pick up the effect of paired distributions over an area and might give a false impression of clustering.
- Scale is important. Depending on the size of the area chosen patterns may be picked up falsely (e.g. if the area is very small a regular pattern may be detected but if the area is expanded a different pattern and result may be obtained).
- Although Rn may suggest a random pattern it may be that the settlements are being affected by relief or soil types which are randomly distributed. In this case these are controlling factors so the settlement pattern is not really random at all.
- Need to have a minimum of 30 points to have statistically valid results (99% significance level).
Nearest Neighbour Analysis
No replies to this topic
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users