Improve Surface Temperature Assimilation in Complex Terrain Areas for the Three-Dimensional Real-Time Mesoscale Analysis (RTMA_3D)

Extending the operational Real-Time Mesoscale Analysis (RTMA) to three dimensions for whole-atmospheric situational awareness and analysis of record is an ongoing project supported in part by the Joint Technology Transfer Initiative (JTTI) and represents a collaboration between scientists from ESRL/GSD and NCEP/EMC. The major goal of RTMA_3D is to: 1) improve tools for situational awareness and nowcasting, 2) provide a three-dimensional analysis of record (AOR) suitable for verification and bias-correction, and 3) accelerate improvement of numerical weather prediction (NWP) models. The new product suite will ultimately improve forecast guidance provided to the public by NWS forecast offices and national centers.

One of the major challenges in RTMA_3D is to generate a fine-resolution and physically coherent 2m temperature analysis, especially in areas of complex terrain. The major observation sources for surface temperature are from MESONET and METAR sites. While METAR observations are generally in good quality and have accurate metadata, the MESONET observations -often crowdsourced – require additional quality control and often lack suitable metadata. For example, it is not uncommon to find MESONET sites with incorrect elevation records, such data require special treatment before further processing and eventual assimilation. In addition, height differences between model and observation elevation often arise and in such situations an adjustment is required. This work will evaluate the impact of adjusting the temperature observation using a local lapse rate prior to assimilation in the RTMA_3D. After a temperature analysis is generated, a follow-up downscaling process similar as “smartinit” is run to get a more-representative surface temperature field for complex terrain areas. The operational “smartinit” package is revisited for the origin of the spurious freckle problem and some improvements have been proposed. A case study of the Colorado mountain area around Gerogetown-Berthound Pass will be presented at the conference.