Wednesday, 10 January 2018: 9:15 AM
Salon F (Hilton) (Austin, Texas)
Temperature inversions in the lower troposphere are a nearly ubiquitous feature in the Arctic.
Inversions form as a result of a number of interacting processes, including warm air advection, surface radiative imbalance, and subsidence. Inversion strength (difference in temperature from inversion base to inversion top) and depth (thickness of the inversion layer) have been found to play an important role in the negative longwave feedback, low-level cloud formation, depth of the mixed layer, and the transport of heat and moisture from openings in the sea ice. We computed inversion strength, depth, base height and frequency for the region north of 65 latitude for all 42 model runs available in the CESM Large Ensemble Community Project. Simulations in the CESM Large Ensemble were performed with the nominal 1-degree latitude/longitude version of the Community Earth System Model version 1, using CAM5.2 as its atmospheric component. Hereafter “CESM” will refer to the 42-member ensemble. In order to capture vertical variability, we used the instantaneous 6-hourly output; for purposes of comparing to the observational record, our analysis is restricted to the period 1990-2005. Model output was compared to twice-daily radiosonde records from 32 Arctic weather stations retrieved from the University of Wyoming archive. Inversions were identified from temperature profiles following the algorithm in Kahl (1990). We discuss our results in the context of previous studies evaluating seasonality and spatial distribution of inversion characteristics from satellite and reanalysis products. Broad seasonal characteristics of inversions in CESM and in the radiosonde observations are presented in Figure 1. Spatial distribution and seasonal changes in surface-base inversion (SBI) frequency compare very well to ERA-Interim data (as computed by Zhang et al 2011), with SBI frequency being highest over land surfaces in the winter and fall. SBIs in the CESM ensemble are more frequent over ice than over open water except during the melt season. Elevated inversions dominate over the Arctic ocean during summer. A seasonal cycle of higher median inversion base heights during summer and lower base heights during winter is apparent over the entire study region with the exception of the North Atlantic ocean. The largest annual changes in median base height occur over ice-free land surfaces. The strongest and deepest inversions in CESM form during winter over the Canadian archipelago and Siberia. Although the season cycle and spatial characteristics are qualitatively accurate in the model, compared to radiosonde observations, inversions in the CESM ensemble are generally biased high in frequency and strength. Elevated inversions in CESM tend to be deeper and lower than in the observations, while SBIs tend to be too shallow in the ensemble. Figure 2 summarizes some findings based solely on the radiosonde data. Using the robust rank-order test described in Lazante (1996), we tested the significance of differences between seasonal medians of inversion depth, base height, and strength for the periods 1990-2000 and 2005-2017. We find there are differences significant at the 99% level in most inversion properties. Most striking is the nearly uniform decrease in median inversion depth. Inversion strength and base height exhibit varying trends, with some stations showing significant decreases and some showing significant increases. We discuss potential causes of changes and potential pitfalls in drawing conclusions from the radiosonde data alone.
Inversions form as a result of a number of interacting processes, including warm air advection, surface radiative imbalance, and subsidence. Inversion strength (difference in temperature from inversion base to inversion top) and depth (thickness of the inversion layer) have been found to play an important role in the negative longwave feedback, low-level cloud formation, depth of the mixed layer, and the transport of heat and moisture from openings in the sea ice. We computed inversion strength, depth, base height and frequency for the region north of 65 latitude for all 42 model runs available in the CESM Large Ensemble Community Project. Simulations in the CESM Large Ensemble were performed with the nominal 1-degree latitude/longitude version of the Community Earth System Model version 1, using CAM5.2 as its atmospheric component. Hereafter “CESM” will refer to the 42-member ensemble. In order to capture vertical variability, we used the instantaneous 6-hourly output; for purposes of comparing to the observational record, our analysis is restricted to the period 1990-2005. Model output was compared to twice-daily radiosonde records from 32 Arctic weather stations retrieved from the University of Wyoming archive. Inversions were identified from temperature profiles following the algorithm in Kahl (1990). We discuss our results in the context of previous studies evaluating seasonality and spatial distribution of inversion characteristics from satellite and reanalysis products. Broad seasonal characteristics of inversions in CESM and in the radiosonde observations are presented in Figure 1. Spatial distribution and seasonal changes in surface-base inversion (SBI) frequency compare very well to ERA-Interim data (as computed by Zhang et al 2011), with SBI frequency being highest over land surfaces in the winter and fall. SBIs in the CESM ensemble are more frequent over ice than over open water except during the melt season. Elevated inversions dominate over the Arctic ocean during summer. A seasonal cycle of higher median inversion base heights during summer and lower base heights during winter is apparent over the entire study region with the exception of the North Atlantic ocean. The largest annual changes in median base height occur over ice-free land surfaces. The strongest and deepest inversions in CESM form during winter over the Canadian archipelago and Siberia. Although the season cycle and spatial characteristics are qualitatively accurate in the model, compared to radiosonde observations, inversions in the CESM ensemble are generally biased high in frequency and strength. Elevated inversions in CESM tend to be deeper and lower than in the observations, while SBIs tend to be too shallow in the ensemble. Figure 2 summarizes some findings based solely on the radiosonde data. Using the robust rank-order test described in Lazante (1996), we tested the significance of differences between seasonal medians of inversion depth, base height, and strength for the periods 1990-2000 and 2005-2017. We find there are differences significant at the 99% level in most inversion properties. Most striking is the nearly uniform decrease in median inversion depth. Inversion strength and base height exhibit varying trends, with some stations showing significant decreases and some showing significant increases. We discuss potential causes of changes and potential pitfalls in drawing conclusions from the radiosonde data alone.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner