WE PROPOSE TO RETRIEVE SURFACE-TO-DEPTH PROFILES OF SOIL MOISTURE FROM MULTITEMPORAL CYGNSS DATA BY UTILIZING OUR RECENTLY DEVELOPED BISTATIC RADAR SCATTERING MODELS. WE WILL ACCOMPLISH THIS FOR BOTH BARE SURFACES AND VEGETATED LANDSCAPES. WE WILL ALSO DEVELOP A METHOD FOR RETRIEVING SOIL MOISTURE AND ACTIVE LAYER THICKNESS FOR PERMAFROST SOILS USING GPS REFLECTOMETRY DATA FOR POSSIBLE FUTURE DATA COLLECTION AT HIGHER LATITUDES AND OVER THE BOREAL AND ARCTIC ZONES. THE RADAR SCATTERING MODELS HAVE SEVERAL UNKNOWNS WHICH WE CAN WITH ASSUMPTIONS AND SIMPLIFICATIONS REDUCE TO JUST 3 OR 4 UNKNOWNS. HOWEVER EVEN USING THREE UNKNOWNS TO FORMULATE THE INVERSE PROBLEM WE ENCOUNTER NON-UNIQUENESS BECAUSE EACH LEVEL-1B CYGNSS OBSERVATION HAS ONLY ONE POLARIZATION CHANNEL NAMELY LEFT-HAND CIRCULAR POLARIZATION (LHCP) AND THEREFORE ESSENTIALLY ONLY ONE DATA POINT FOR AN OBSERVATION CELL. THE PROBLEM IS FURTHER EXACERBATED IF VEGETATION PARAMETERS ARE CONSIDERED UNKNOWNS. WE THEREFORE PROPOSE TO BUILD A DATA SPACE OF LARGER DIMENSIONALITY BY ACCUMULATING MULTIPLE OBSERVATIONS WITHIN REASONABLE SPATIOTEMPORAL BOUNDS: FOR ANY GIVEN DDM PIXEL WITHIN THE CYGNSS COVERAGE AREA WE WILL UTILIZE THE LEVEL-1B SCATTERING CROSS SECTION DATA BUT WE WILL ACCUMULATE AN ARRAY OF THESE DATA FROM AS MANY OBSERVATIONS AS POSSIBLE WITHIN A 24-HOUR WINDOW. THESE OBSERVATIONS WILL PROVIDE DIVERSITY IN INCIDENCE ANGLES DURING PERIODS OF TIME WHEN SOIL MOISTURE IS EXPECTED TO BE CONSTANT TO WITHIN THE RETRIEVAL UNCERTAINTY. ONCE THE PLURALITY OF DATA POINTS IS ACCUMULATED WE WILL SOLVE THE SOIL MOISTURE PROFILE INVERSION PROBLEM VIA AN ITERATIVE NONLINEAR OPTIMIZATION METHOD. OUR BISTATIC RADAR SCATTERING MODEL WILL BE THE CORE OF THE ITERATIVE ALGORITHM. WE WILL INCLUDE PROVISIONS FOR SUBSURFACE SOIL MOISTURE PROFILES IN BOTH THE FORWARD MODEL (LAYERED RANDOM ROUGH SURFACES) AND THE INVERSE MODEL (QUADRATIC PROFILE WITH 3 UNKNOWN COEFFICIENTS). A COMPREHENSIVE SET OF FORWARD MODEL SIMULATIONS WILL BE PERFORMED TO UNDERSTAND THE SENSITIVITIES OF THE OBSERVED LEVEL-1B PRODUCTS TO VARIOUS LANDSCAPE PARAMETERS AND TO ASSESS MODEL ACCURACY VS. CYGNSS OBSERVATIONS. ADDITIONALLY A THOROUGH UNCERTAINTY ANALYSIS WILL BE CARRIED OUT TO QUANTIFY THE EXPECTED RETRIEVAL ERRORS AS A FUNCTION OF FORWARD MODEL AND PARAMETERIZATION ERRORS OBSERVATION (INCLUDING CALIBRATION) ERRORS AND OPTIMIZATION ERRORS THE PROPOSED PROJECT WILL DELIVER RZSM LEVEL-2 PRODUCTS WITH QUANTIFIED UNCERTAINTIES AND AT DAILY TEMPORAL SCALES. THE SPATIAL RESOLUTION WILL BE COMMENSURATE WITH THE MISSIONS OBSERVATIONS GEOMETRY AND WILL BE AS CLOSE AS POSSIBLE TO THE SMALLEST CELL SIZE AT THE SPECULAR POINT. THE MAPS WILL INCLUDE AREAS REPRESENTING DIVERSE LANDCOVER TYPES WHICH ARE PRESENT IN THE +/-38 DEGREE LATITUDE OF CYGNSS COVERAGE DOMAIN. SINCE NO CYGNSS OBSERVATIONS ARE AVAILABLE IN PERMAFROST LANDSCAPES ANALYSES FOR SUCH REGIONS WILL FOCUS ON SIMULATIONS.