A Community Retrieval for Multi-Instrument Thermodynamic Profiling of the Boundary Layer
Authors
Jonathan Gero — University of Wisconsin
David D. Turner — NOAA- Global Systems Laboratory
Tim Wagner — University of Wisconsin, Madison
Raymond Garcia — University of Wisconsin
Denny John Hackel — University of Wisconsin
Category
Boundary layer structure, including land-atmosphere interactions and turbulence
Description
The AERIoe algorithm produces accurate retrievals of temperature and water vapor profiles in the boundary layer (BL). It operates in all-sky conditions up to the cloud-base height. It was developed to use AERI observations, but it can seamlessly incorporate additional instrument data streams to increase accuracy, e.g. surface MET, surface remote sensors (microwave radiometer, lidar, etc.), aircraft observations, satellite observations, model data. The algorithm includes an uncertainty estimate for each vertical profile. The retrievals have been rigorously validated in various climactic regimes through comparisons with radiosonde observations.
An impediment to widespread adoption of AERIoe by the broader scientific community is that it is written in a proprietary programming language (IDL), which hinders both the distribution and future development of the retrieval. In order to address this problem, we are planning to convert AERIoe into an open-source community retrieval algorithm that any investigator can run for free, and contribute to testing and improving the algorithm. The new retrieval algorithm will also be more flexible and modular, to make it simple to use additional radiative transfer models and input instrument datastreams. Free availability of this algorithm will facilitate increased adoption of BL profiling data in the community, the incorporation of ground-based BL observations into NWP models, leading to improved forecasts, in particular for high-impact weather events. It can also spur the deployment of future networks of instruments, further augmenting forecasting capabilities.
The existing AERIoe algorithm will be converted into the open source Python language, and we will create a platform-independent installation package. The code base will be posted publically to GitHub. Investigators from the community will have the opportunity to download, test and make improvements to the algorithm. The conversion to Python will enable easy integration with the widely used SHARPpy analysis package, such that derived quantities like CAPE and CIN (and their uncertainties) can be calculated along with the retrieved profiles.
