PROCESSING SATELLITE ALTIMETRY DATA REQUIRES THE COMPUTATION OF PATH DELAY IN THE NEUTRAL ATMOSPHERE THAT IS USED FOR CORRECTING RANGES. THE PATH DELAY IS COMPUTED USING NUMERICAL WEATHER MODELS AND THE ACCURACY OF ITS COMPUTATION DEPENDS ON THE ACCURACY OF NUMERICAL WEATHER MODELS. ACCURACY OF NUMERICAL MODELS OVER LAND CAN BE EVALUATED BY COMPARISON WITH GROUND METEOROLOGICAL STATIONS ATMOSPHERIC SOUNDINGS AND WITH ESTIMATES OF PATH DELAY FROM SPACE GEODESY OBSERVATIONS. HOWEVER THE ACCURACY OF NUMERICAL WEATHER MODELS OVER ANTARCTICA AND GREENLAND WHERE THERE IS A VERY SPARSE NETWORK OF GROUND STATIONS IS NOT WELL KNOWN. THEREFORE THE ACCURACY OF THE PATH DELAY COMPUTATION OVER THESE REGIONS IS NOT KNOWN. THE ESSENCE OF THIS PROPOSAL IS TO USE GPS RADIO OCCULTATION DATA FOR COMPARISON OF OBSERVED PATH DELAYS FROM THESE OBSERVATIONS WITH PREDICTED PATH DELAYS FROM THE NUMERICAL WEATHER MODELS THAT ARE USED FOR PROCESSING ICESAT DATA. THE DIFFERENCES WILL BE USED FOR DERIVING THE ERROR MODEL OF ATMOSPHERIC PATH DELAY CORRECTIONS OF ALTIMETRY DATA OVER ANTARCTICA AND GREENLAND. SINCE THERE ARE TOO FEW METEOROLOGICAL STATIONS IN THE INNER PARTS OF ANTARCTICA AND GREENLAND THE GPS RADIO OCCULTATIONS WILL BE USED AS THE GROUND TRUTH. THE NEAR TERM GOAL IS TO EVALUATE THE CONTRIBUTION OF THE UNCERTAINTY OF OUR KNOWLEDGE OF THE STATE OF THE ATMOSPHERE TO ESTIMATES OF ICE MASS LOSS IN ANTARCTICA AND GREENLAND AND ITS SEASONAL VARIATIONS. IN PARTICULAR WE WILL EVALUATE SYSTEMATIC ERRORS IN PATH DELAY SEASONAL VARIATIONS AND THE SECULAR TREND. SEASONAL ERRORS MAY CAUSE BIASES IN ESTIMATES OF ICE MASS BALANCE DERIVED FROM SATELLITE ALTIMETRY. AS A LONG TERM GOAL WE WILL EVALUATE THE FEASIBILITY OF USING RADIO OCCULTATION DATA NOT ONLY FOR THE CHARACTERIZATION OF PATH DELAY ERRORS BUT FOR AN IMPROVEMENT IN COMPUTATION OF PATH DELAY FOR PROCESSING ICESAT-2 DATA. WE WILL DERIVE AN EMPIRICAL MODEL OF BIASES IN PATH DELAYS AND EVALUATE WHETHER THE BIAS CORRECTION MERITS PROCESSING ICESAT-2 DATA OVER ANTARCTICA AND GREENLAND. IN A CASE IF THE EVIDENCE WILL BE COLLECTED IN FAVOR OF THIS MODEL THE EXPERIENCE GAINED IN PROCESSING RO DATA WILL HELP TO FORMULATE RECOMMENDATIONS FOR MODIFICATION OF THE DATA ANALYSIS PROCEDURE TAKING INTO ACCOUNT ALL ISSUES OF LOGISTICS.