The use of radio spectrum is well established for meteorological and climate observations. A number of discrete bands (centred on 1.413, 6.9, 10.7, 18.7, 23.8, 31.4, 36.5, 50.2-57.3, 89, 157, 166, 175-192 GHz) already provide unique information on soil moisture, sea and land surface temperature, atmospheric temperature, atmospheric water (in vapour, liquid and ice phases, including precipitation), marine near surface wind vectors, snow on the ground, sea ice and vegetation parameters. These observations are critical both the weather forecasts using Numerical Weather Prediction (NWP) on High Performance Computers, and for climate monitoring. In NWP passive and actively sensed observations using these bands provide over half of forecast error reduction. In the near future a number of bands will also provide more detailed information on ice clouds using bands above 200 GHz. There is a direct link between the availability of RFI-free spectrum in these bands and the major socio-economic benefits of NWP. The data assimilation systems used to initialise NWP weather forecasts are also now used for reanalysis of historical observations over long time periods, in a process known as reanalysis. Reanalysis therefore also depends critically on RFI-free spectrum in these bands and if RFI levels increase in future this may risk introducing spurious trends. RFI is already being detected in bands below 20 GHz, and there is particular concern over the lack of an allocation to Earth observation at C-band for passive ocean monitoring.