Suppose you have four temperature stations in a relatively small geographical area, and you add a fifth in the geographic middle of the group. Should the trend change?
No, it shouldn’t. The addition of the fifth station might cause the average temperature of the group to shift due to differences in local geography, but it shouldn’t affect the trend much. This is because the temperature trend in the geographic middle of the group should be about the same as the numerical average of the trend at the other four stations.