Date of Award


Document Type


Degree Name

Master of Science in Coastal Marine and Wetland Studies


Coastal and Marine Systems Science

First Advisor

Erin E. Hackett

Second Advisor

Var Limpasuvan

Third Advisor

Richard F. Viso


Radar is a remote sensor that is useful in scientific and military applications. The environment affects the accuracy of radar measurements as well as the predictability of a radar system’s performance. Because of the complexity of the dynamic processes occurring in the marine atmospheric boundary layer (MABL), which includes the lowermost troposphere and ocean surface, the impact of the environment on radar is intricate and difficult to assess. To better understand the relative importance of various aspects of the MABL environment on radar wave propagation, this study evaluates the sensitivity of radar wave propagation to the MABL environment using a global sensitivity analysis (SA) method, the extended Fourier amplitude sensitivity test (EFAST), and the Variable Terrain Radio Parabolic Equation (VTRPE) simulation, which calculates propagation power of radar waves in a wide variety of marine atmospheric conditions. A total of 16 environmental parameters are examined, 8 parameterizing the rough ocean surface, and 8 parameterizing the atmospheric vertical refractivity profiles. Radar frequencies of 3, 9, and 15 GHz are each simulated with horizontal (HH) and vertical (VV) polarization, resulting in sensitivity calculations for 6 different cases. The study is conducted for a domain of 1 km in altitude and 60 km in range using a low grazing angle generic air/sea surveillance radar. The relative importance of the different parameters varied much more with frequency than polarization. The EFAST method takes into account parameter interactions, which are found to be significant and can be essential to correctly interpret the significance of a parameter. Results show that the atmospheric mixed layer parameters are most important, particularly the height of the mixed layer. Overall, swell period is the most significant ocean surface parameter. However, sea directionality is also important at 3 GHz, and sea surface roughness and salinity are important at 9 and 15 GHz, respectively. Sensitivities to ocean surface parameters, except those related to directionality, become more prominent as radar frequency increases, and some sensitivity differences with respect to polarization occur regarding sea surface characteristics. Due to spatial variability of sensitivity throughout the domain, regional analysis is performed, using short (0-10 km), mid (10-30 km), and long (30-60 km) range, and low (0-200 m), mid (200-600 m), and high (600-1000 m) altitude divisions (9 regions). The most sensitive parameter in each low altitude region, from short to long range, is evaporation duct height and mixed layer height (mid and long range). The mixed layer height is the most sensitive parameter in all mid-altitude regions. At high altitude, the most sensitive parameter varies with frequency, except at short range where it is the mixed layer refractivity gradient (i.e., M-gradient). At mid-range, the most sensitive parameters are the inversion layer strength, mixed layer M-gradient, and mixed layer height for 3, 9, and 15 GHz respectively. At long range, the inversion strength is the most sensitive parameter at 3 GHz, while at 9 and 15 GHz it is the wind speed. These regional sensitivity results, along with those for the whole domain, can be used to determine which environmental parameters need to be specified with high accuracy when accounting for their effects on propagation for various radar systems and applications. This sensitivity information can also be used to help guide field measurements for simulation validation studies as it indicates what aspects of the environment need to be focused on for such experimental campaigns. Furthermore, these results provide guidance on prioritization of environmental characterization in numerical weather prediction (NWP) and inversion studies (e.g., refractivity from clutter (RFC) studies), which are the two most common numerical methods currently used to address environmental effects on propagation. Additionally, the methodology presented in this study can be used and applied to similar problems that seek to understand the sensitivity to environmental effects on other remote sensors, such as infrared (IR), optical, and acoustic sensors.