report/ab35.html

---
layout: default
title: Report Series
---

<html>


<head>
   <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
   <meta name="Author" content="Anna McCravy">
   <meta name="GENERATOR" content="Mozilla/4.51 [en] (WinNT; U) [Netscape]">
   <title>rpt35</title>
</head>
<body text="#000000" bgcolor="#FFFFFF" link="#0000EE" vlink="#551A8B" alink="#FF0000">
<h1>Report 35:&nbsp; Structure of the Tropical Lower Stratosphere
as Revealed by Three Reanalysis Data Sets</h1>
<ul><b>Pawson, </b>Steven and Michael Fiorino
<br>May 1996, 35 pp.&nbsp;
<hr>While the skill of climate simulation models has advanced over the
last decade, mainly through improvements in modeling, further progress
will depend on the availability and the quality of comprehensive validation
data sets covering long time periods. A new source of such validation data
is atmospheric "reanalysis" where a fixed, state-of-the-art global atmospheric
model/data assimilation system is run through archived and recovered observations
to produce a consistent set of atmospheric analyses. Although reanalysis
will be free of non-physical variability caused by changes in the models
and/or the assimilation procedure, it is necessary to assess its quality.
That is, how good is reanalysis and how close should our climate models
agree with these data.
<p>A region for stringent testing of the quality of reanalysis is the tropical
lower stratosphere. This portion of the atmosphere is sparse in observations
but displays the prominent quasi-biennial oscillation (QBO) and an annual
cycle, neither of which is fully understood, but which are likely coupled
dynamically. We first consider the performance of three reanalyses, from
NCEP/NCAR, NASA and ECMWF, against rawinsonde data in depicting the QBO
and then examine the structure of the tropical lower stratosphere in NCEP
and ECMWF data sets in detail.
<p>While the annual cycle and the QBO in wind and temperature are quite
successfully represented, the mean meridional circulations in NCEP and
ECMWF data sets contain unusual features which may be due to the assimilation
process rather than being physically based. Further, the models capture
the long-term temperature fluctuations associated with volcanic eruptions,
even though the physical mechanisms are not included, thus implying that
the model does not mask prominent stratospheric signals in the observational
data. We conclude that reanalysis offers a unique opportunity to better
understand the dynamics of QBO and can be applied to climate model validation.
<i><a href="pdf/35.pdf">(pdf
file)</a></i>
</ul>

<p><font size=-1>UCRL-MI-123395</font></p>