Using Observing Systems to Differentiate Climate Models According by Their Low-Cloud Feedback Strengths

Satellite-based remote sensing observations may be able to provide constraints on important and poorly-understood climate system low-cloud feedbacks. To evaluate remote sensing system specifications for providing such constraints, we present a hyperspectral shortwave and infrared observing system simulation experiment (OSSE) that produces simulated top-of-atmosphere measurements and broadband OLR and albedo derived from climate model fields. The OSSE has already been validated against published standards and observations, and we use it to compare and contrast observational signatures of climate change from different instances of the NCAR CCSM3 model and GFDL CM2 model. The models exhibit intra- and inter-model differences in low-cloud feedback strengths, even with identical initial conditions and forcings. We analyze these observational signatures to determine the characteristics of an observational record that would be best suited for distinguishing these models based on their feedback strengths.

We find that nearly two decades of continuous, stable hyperspectral measurements are needed to differentiate models where cloud-feedbacks differ by 0.3 W/m²/ºK given significant inter annual variability. Three decades of continuous, stable OLR and albedo measurements is required to achieve a similar level of model differentiation. This finding suggests that proposed measurement systems such as CLARREO or TRUTHS have planned specifications that will establish a measurement baseline that can be used provide observational constraints on climate model feedbacks.