Analysis of Temperature and Humidity Sounding Using Compact Mid-Wave Infrared Instruments

Humidity and temperature profiles produced from infrared (IR) soundings of the atmosphere are vital inputs for our weather forecasting capabilities. Typically, IR sounders measure the IR water band centered around 6.3-μm for humidity profile retrievals, while atmospheric temperature retrievals are generated in part using the 15-μm CO₂ bands. Because the 15-μm CO₂ band is relatively broad, it facilitates vertical sampling of the atmosphere with a moderate spectral resolution instrument. However, the requirement to measure out to 15-μm in the long-wave infrared (LWIR) requires a sensor with a large optical aperture to meet horizontal footprint requirements and a very cold detector (40-60K) for appropriate sensitivity. These two requirements drive up the size and complexity of the instrument.

In the study presented here, we explore the potential of sounding exclusively in the mid-wave infrared (MWIR) using the blue end of the water band and the 4.3-μm CO₂ line for humidity and temperature profile measurements. While the 4.3-μm band is significantly narrower than the 15-μm feature, the use of high spectral resolution spectrometers that make use of new large-area, digital focal plane arrays (FPA) allow details of the 4.3-μm feature to be resolved. The improved noise characteristics of these digital FPAs improve the signal to noise ratio of the measurements; the large area (i.e. larger number of pixels) allows for finer spectral binning.

To perform this study, we put together a matched suite of tools which can be used to evaluate MWIR sounder concepts. We use LBLRTM as a forward model to simulate upwelling radiation and compute relevant Jacobians. We use a complete signal chain analysis for a dispersive spectrometer to simulate the noise characteristics using sounder design parameters and collection conditions as inputs. Finally, we perform full simulated retrievals using an algorithm similar to that used by an established sounder, CrIS. With these tools, we are able to perform a complete end-to-end simulation of multiple sounder concepts and evaluate how final data products vary with system design parameters.

We have used these tools to study how sounder spectral coverage, spectral resolution, and sensor noise characteristics impact temperature and humidity retrieval accuracy and vertical resolution. We determine the vertical resolution by studying the information density of the signal. To determine retrieval accuracy, we first perform full sounder simulation for a representative set of profiles taken from the CIMSS atmospheric retrieval training database. We then compare retrieved profiles to the truth.