VALIDATING THE BANDFIELD SURFACE ROUGHNESS MODEL USING NOVEL THEMIS ROTO DATA: APOLLINARIS MONS IN A NEW LIGHT

New THEMIS routine off-nadir targeted observation (ROTO) data provide a novel approach to quantify centimeter-scale surface roughness of Mars and provide the means to test prior roughness estimates by Josh Bandfield in his 2009 modeling paper. Here, we examine the effects of sub-pixel surface roughness on the derived THEMIS brightness temperature (BT). Rough surfaces with sub-pixel topographic slopes typically produce anisothermal pixel-integrated surface temperatures due to changes in the degree of shadowing at varying solar incidence angles. This anisothermality results in a non-uniform BT with higher temperatures at shorter wavelengths. Subsequent separation of temperature and emissivity is negatively impacted, hindering emissvity-based compositional analysis. Due to the sensitivity of TIR spectra to this scale of roughness, directional emission measurements can be used to retrieve macro-scale surface roughness. Surface shadowing produces a larger range of temperature variation with respect to wavelength and emission angle.

A series of ROTO data centered around Apollinaris Mons were acquired at local true solar times from 18:00h -19:00h, solar longitudes from 38°- 47°, incidence angles from 94°-102°, and emission angles from -31°to 33°. Using these eight observations, BT differences were examined. Observations show temperature asymmetry about the -2 roll angle (effectively nadir) with the lowest temperatures recorded at the highest magnitude emission angles. Negative roll angles, highlighting western facing slopes, show higher overall temperatures than the positive roll angles of similar magnitude. This corresponds to warmer western facing slopes that had previously been illuminated by the setting sun. BT variations between THEMIS bands 3 and 9 indicate the range of temperatures within the region of study. Band 3 BT are always greater than band 9 BT, indicating the presence of sub-pixel anisothermality.

By combining the KRC thermal model and a modified Bandfield surface roughness model, simulated BT data are used to quantify this sub-pixel anisothermality. Maximizing emission angle observations with variable look angles provides a more complete view of surface features, improving previous estimates of RMS and Θ-bar roughness by Bandfield (2009).