Ozone sensitivity of tropical upper‐troposphere and stratosphere temperature in the MetOffice Unified Model
A systematic warm bias in the tropical tropopause layer (TTL) is commonly found in both climate and numerical weather prediction models. In this study, the nature of this temperature bias is examined by integrating the MetOffice Unified Model (MetUM) with various ozone concentrations in the TTL. Lik...
Gespeichert in:
Veröffentlicht in: | Quarterly journal of the Royal Meteorological Society 2018-07, Vol.144 (715), p.2001-2009 |
---|---|
Hauptverfasser: | , , , , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | A systematic warm bias in the tropical tropopause layer (TTL) is commonly found in both climate and numerical weather prediction models. In this study, the nature of this temperature bias is examined by integrating the MetOffice Unified Model (MetUM) with various ozone concentrations in the TTL. Like other models, the long‐term integration of MetUM with the Atmospheric Model Intercomparison Project (AMIP) configuration shows a notable warm bias (∼2 K) in the TTL with a comparable cold bias in the tropical stratosphere above ∼70 hPa. We demonstrate that these biases are particularly sensitive to the tropical ozone concentration prescribed in the model. By replacing the background ozone, which is typically used for AMIP‐type simulations, with the Southern Hemisphere ADditional OZonesondes (SHADOZ) measurements or the Binary Data Base of Profiles (BDBP), the dipolar temperature biases in the TTL and tropical stratosphere are significantly reduced. Further sensitivity tests show that the tropical ozone amount in a 14–20 km layer is a key contributor to this change. These results suggest that accurate ozone forcing in the TTL is crucial for reliable weather and climate simulations.
Latitude–pressure distributions of zonal‐mean temperature bias for the 20‐year (1989–2008) AMIP simulations: (a) CTR, (b) SOZ, (c) BDBP, (d) SOZ_TTL and (e) BDBP_TTL. The biases are computed from ERAI temperatures for the same period. Shading and contour intervals are 0.2 and 0.5 K, respectively. |
---|---|
ISSN: | 0035-9009 1477-870X |
DOI: | 10.1002/qj.3346 |