I have been running the numbers for potential interference levels at Boscombe Down on 2.3GHz from this location. Using the G0MJW software and the simplified path model ( no terrain data) it would appear that my 100W at the feed point and 20dBi antenna ( possibly 22dBi) gives around 6dB margin over what would be required. These numbers are for a 10% exceed.
The overall loss is 205dB, dominated by tropo scatter with diffraction and refraction both being significantly higher.
The G3YGF plots in Vol 1 of the RSGB Microwave Handbook confirm these losses.
It looks like I can operate without being too concerned that I am causing interference to the MOD down in BD!