Mid-Course Space Experiment (MSX) infrared (IR) observations in the earth limb were used to obtain spatial power
spectral densities (PSDs) for five sensor bands over a wide range of earth limb background clutter conditions. These
backgrounds include daytime, nighttime, terminator, aurora, polar mesospheric cloud, atmospheric gravity wave,
stratospheric warming, airglow, and other observations collected over approximately 100 episodic data collection events.
Using a subset of detectors and restricting detector tangent altitude variations, a total of more than 33,000 high-quality
PSDs were generated. For infrared detection of unresolved objects where the solid angle of the object is much smaller
than the instantaneous field-of-view of a sensor element, the spectral component at high spatial frequencies is a critical
metric. PSDs were therefore constructed in the spatial domain using one minute data segments, which allowed spatial
scale assessment from 0.01-10 cycles/km. PSDs that met the clutter model selection criteria were identified,
accumulated, and processed to obtain a small set of empirical, altitude-based model parameters. We describe the MSX
sensor bands, data and data processing employed for PSD generation and final reduction to obtain model parameters.
Key model features are discussed with emphasis on object detection against stressing limb backgrounds. The model was
constructed in a way that facilitates optical design and system engineering application. In particular, it may be used to
address Space Situational Awareness (SSA) questions.