Observation, Prediction, and Modeling Atmospheric Structure Effects on EO/IR Systems

EO/IR sensors observing the battlespace environment through the earth's atmosphere can be adversely affected by spatial and temporal variations in atmospheric radiance and transmission along the sensor line of sight (LOS). The physics of stochastic fluctuations is largely understood, and the radiation transport theory and models that include stochastic effects exhibit high fidelity when compared to corresponding sensor measurements. Deterministic structure is not as well understood, however, and the associated radiance levels and variability are often significantly higher than those of the benign, stochastic background. Since the radiance measured by the sensor comes from both the object of interest and the radiating atmosphere and it can also be attenuated by the atmosphere along the optical path, atmospheric structure and clutter affect target acquisition, identification, discrimination, and tracking in ways that are difficult to assess.

We ranked MSX SPIRIT III radiometer measurements by radiance and clutter levels for a number of altitudes and sensor bands, and the atmospheric phenomena responsible for elevated levels were identified. Many of these high altitude atmospheric structures are not included in current IR radiance codes, and we describe recent efforts to extract and characterize these features using models such as MODTRAN, SHARC, and SAMM2. Our process involves identifying scenes with suitable structure, determining ambient model background conditions for a sufficiently large number of runs where the sensor, line of sight, and geophysical conditions are duplicated, and extracting radiance enhancements due to this structure on a pixel by pixel basis. We present a number of examples where extraction techniques have been successfully applied to scenes from the MSX SPIRIT III radiometer, including phenomena such as aurora, polar mesospheric clouds (PMC), and stratospheric warmings. The extracted structure features are then recombined with new ambient model background scenes such that they are properly located in the global/geophysical environment, accounting for the dependence of specific types of atmospheric structure on time, latitude, and season, in order to ensure real world fidelity.

There are several limitations to this approach, so we used existing model capabilities to address these limitations. In the case of aurora, we show how auroral observations in the infrared by the MSX SPIRIT III radiometer can be used to determine valid model inputs. In the case of stratospheric warmings, we use measured scenes to determine stratospheric temperature enhancements. We demonstrate that proper combination of validated model inputs allows simulation of complex scenes in a real world context, and that prediction can be extended to other IR bands.

The goal of this effort is to develop real time nowcast and forecast capability to estimate EO/IR sensor impairment levels on SSA systems due to geophysical effects on atmospheric structure, and we will discuss plans to develop real time data assimilation capabilities to support operational application.