In this article, we use the analytical Kirchhoff solution (AKS) and numerical Kirchhoff approach to study the bistatic scattering field ($gamma $) from mountain terrain at P-band frequency. The study area is Grand Mesa, Colorado, USA, and the properties of land surface roughness are extracted from airborne lidar surveys. The bistatic scattering coefficient $gamma $ of variance fields, denoted by${gamma }_v$, for several cases of radar resolutions over a 3.6 km by 3.6 km area are calculated at various scattering azimuth angles. Based on the lidar measurements, the land surface is decomposed into ${f}_2 + {f}_3$, where ${f}_3$ is 30 m of deterministic planar patches to approximate the coarse topography and ${f}_2$ is modeled by random rough surfaces with correlation functions. Surface roughness statistics derived from the Lidar data give a typical root mean square height of 0.07 m and a correlation length of 3.6 m for ${f}_2$. The mean values of slopes of ${f}_3$ are 1.3° and 0° with a standard deviation of 1° each, respectively in the x and y directions. Simulations using AKS show that the values of bistatic scattering coefficients for the variance of scattered fields can reach above 10 dB over a range of azimuth angles ${phi }_s$ in the vicinity of the specular direction. Even in mountainous regions, the value of the ${gamma }_v$ around the forward scattering direction is much larger than that for radar backscattering, and thus could support the use of a synthetic aperture radar concept based on signals of opportunity with data acquisition near the forward direction.In this article, we use the analytical Kirchhoff solution (AKS) and numerical Kirchhoff approach to study the bistatic scattering field ($gamma $) from mountain terrain at P-band frequency. The study area is Grand Mesa, Colorado, USA, and the properties of land surface roughness are extracted from airborne lidar surveys. The bistatic scattering coefficient $gamma $ of variance fields, denoted by${gamma }_v$, for several cases of radar resolutions over a 3.6 km by 3.6 km area are calculated at various scattering azimuth angles. Based on the lidar measurements, the land surface is decomposed into ${f}_2 + {f}_3$, where ${f}_3$ is 30 m of deterministic planar patches to approximate the coarse topography and ${f}_2$ is modeled by random rough surfaces with correlation functions. Surface roughness statistics derived from the Lidar data give a typical root mean square height of 0.07 m and a correlation length of 3.6 m for ${f}_2$. The mean values of slopes of ${f}_3$ are 1.3° and 0° with a standard deviation of 1° each, respectively in the x and y directions. Simulations using AKS show that the values of bistatic scattering coefficients for the variance of scattered fields can reach above 10 dB over a range of azimuth angles ${phi }_s$ in the vicinity of the specular direction. Even in mountainous regions, the value of the ${gamma }_v$ around the forward scattering direction is much larger than that for radar backscattering, and thus could support the use of a synthetic aperture radar concept based on signals of opportunity with data acquisition near the forward direction. Leer más