Transcript Fu et al. (2004)

Global Warming Seen from
Satellites: A Recent Debate
on Tropospheric
Temperature Trends
Qiang Fu
Dept. of Atmospheric Sciences
University of Washington
Presentation Outline
 Tropospheric
Temperature Versus Surface
Temperature Warming: A Paradox
 MSUs on NOAA Polar Orbiting Satellites
 Stratospheric
Contamination & Correction
 Vertical
Structure of Tropical Tropospheric
Temperature Trends
 Pole-ward
Shift of Tropospheric Jet Streams
and Hadley Circulation Broadening
 Tropospheric
Trend Patterns in the Antarctica
Global Surface Temperature
Variations
IPCC 2007: Summary
for Policymakers
IPCC2001
“It is likely that
there have been
real differences
between the rate of
warming in the
troposphere and
the surface over
the last twenty
years, which are
not fully
understood”
__IPCC (2001).
GCM versus Obs. for Trend
Differences
Santer et al. (2000, Science)
How Can We Explain the Paradox
• Global climate models are missing something
important?
•
•
(e.g., Bengtsson et al. 1999; Santer et al. 2000; 2003; Hegerl
and Wallace 2002; Hansen et al. 2002)
Problems in surface temperature data?
(e.g., Kalnay and Cai 2003; Trenberth 2004; Parker 2004)
Problems in tropospheric temperature data?
(e.g., Seidel et al. 2004; Hurrell and Trenberth 1997; Mears et
al. 2003; Vinnikov and Grody et al. 2003)
The US Climate Change Science Program (CCSP) is preparing more
than 20 synthesis and assessment reports by the end of 2007: The
first topic is temperature trends in the lower atmosphere (April 2006).
Radiosonde Temperatures
Advantages
• Long record (1950s)
• Good vertical resolution
Disadvantages
• Many changes in instruments
and observation methods
• Known and unknown biases
• Sparse coverage
-0.03 to 0.04 K/decade for
1979-2001 (Seidel et al. 2004)
MSU Observations from NOAA
Polar-Orbiting Satellites
• Global coverage
• Data since late of 1978
• All weather conditions
MSU: 4 channels (AMSU:15)
• Channel 2: Mid-troposphere (53.74 GHz)
• Channel 4: Stratosphere (57.95 GHz)
Climate monitoring
(Spencer & Christy 1990)
Satellite Data Analyses
• Satellite local sampling-time drifts
• MSU calibrations (inter-satellites)
• Satellite orbit decays
(e.g., Christy et al. 1995; Wentz et al. 1998; Christy et
al. 1998; Prabhakara et al. 2000; Christy et al. 2000;
Mo et al. 2001; Christy et al. 2003; Mears et al. 2003;
Vinnikov and Grody 2003)
A continuing data-analysis effort has been made to
satisfy climate research requirements of homogeneity
and calibration.
MSU Scan Pattern
T4 = (T44+T45+T46+T47+T48)/5
T2 = (T24+T25+T26+T27+T28)/5
T2LT = (T23+T24+T28+T29)-3(T21+T22+T210+T211)/4
(Spencer and Christy 1992)
Tropospheric Temperature Trends
from MSU (1/1979-12/2001)
•
Univ. of Alabama at Huntsville (UAH)
Mid-troposphere (T2): 0.01K/decade
Low-mid troposphere (T2LT): 0.055K/decade
•
•
(Christy et al. 2003)
Remote Sensing System (RSS)
Mid-troposphere (T2): 0.1K/decade (Mears et al. 2003)
Surface Trend
0.17K/decade (Jones & Moberg 2003)
We argue that the trends reported by both teams for the
“mid-troposphere” channel are substantially smaller than
the actual trend of the mid-tropospheric temperature.
___ Fu et al. (2004)
Satellite Observed Brightness
Temperature

Tb  TsWs   T (z)W (z)dz,
0
where Ts is the surface temperature, Ws the
surface contribution factor, T(z) the atmospheric
temperature profile, and W(z) the weighting
function.
1
Height (km)
Stratospheric Temperature Anomaly (K)
Stratospheric Temperature Anomaly (K)
Weighting Function and Tb Response
1.2
48
(a)
MSU Channel 4
MSU Channel 2
10
Pressure (hPa)
31
Stratosphere
100
(b) MSU2
0.2
0.6
0.1
0
0
-0.1
-0.6
-0.2
-1.2
-0.2
-0.1
0
0.1
0.2
Tropospheric Temperature Anomaly (K)
1.2
(c) MSU4
1
16
Tropopause
Troposphere
1000
0
0.02 0.04 0.06 0.08
0
0.1 0.12
Weighting Function (km-1)
0.6
0
-0.6
0.5
0
-0.5
-1
-1.2
-0.2
-0.1
0
0.1
0.2
Tropospheric Temperature Anomaly (K)
Fu et al. (2004)
Observed Stratospheric Cooling
Ramaswamy et al.
(2001)
Therefore T2 by itself is not a good indicator for the temperature trend in
the troposphere because it reflects combined influences of stratospheric
and tropospheric changes, which largely cancel each other.
Removing Stratospheric Contamination
T2LT created by Spencer and Christy (1992)
[T2LT = (T23+T24+T28+T29)-3(T21+T22+T210+T211)/4]
• Amplify noise by more than an order of magnitude
• Increase inter-satellite calibration biases
• Sensitive to surface variations and mountainous terrain
(e.g., Hurrell & Trenberth 1997; Wentz
& Schabel 1998; Swanson 2003)
Although a stratospheric influence on the T2 trend has
long been recognized, it has never been quantified.
__ Fu et al. (2004)
What is the tropospheric temperature trend
based on satellite MSU observations?
Methodology
• A new approach to remove the stratospheric
contamination by using data from MSU channel 4
• Free of the complications afflicting T2LT
We define the free-tropospheric temperature
as the mean temperature between 850 and
300 hPa (TTR). We derive this temperature
from the measured brightness temperatures of
MSU channels 2 and 4, as
TTR = a0 + a2T2 + a4T4.
__ Fu et al. (2004)
Coefficients a0, a2 & a4 (1)
• Radiosonde data from Lanzante, Klein, Seidel (LKS)
87 stations
15 pressure layers
1000-10 hPa
1958 - 1997
Lanzante et al. (2003)
• Applying the weighting functions to the radiosonde
data to simulate T2 and T4
Global-, hemispheric- and tropical-average
monthly anomalies for TTR, T2, and T4
Time Series of Monthly mean,
global temperature anomalies
Global Temperature Anomalies (K)
1.5
1
0.5
0
-0.5
RSS: T_2
-1
RSS: T_4
RSS: T_850-300
-1.5
197919811983198519871989199119931995199719992001
Year
Fu et al. (2004)
Temperature Trends (1)
0.3
Temperature Trends (K/decade)
(a)
UAH: T_2
0.25
RSS: T_2
0.2
Surface Temp. (4, 5)
0.15
0.1
0.05
0
-0.05
1979-2001
-0.1
Globe
NH
SH
Tropics
0.3
Temperature Trends (K/decade)
(b)
UAH: T_850-300
0.25
RSS: T_850-300
0.2
Surface Temp. (4, 5)
0.15
0.1
0.05
0
-0.05
1979-2001
-0.1
Globe
NH
SH
Tropics
Fu et al. (2004)
Temperature Trends (2)
• The stratospheric contamination in T2 trend is
-0.08 K/decade (Fu et al. 2004).
• Based on RSS MSU data, the ratio of tropospheric
temperature trend to surface temperature trend is
~1.1 for the globe and 1.6 for the tropics (Fu et al.
2004).
• For T2 trends of 0.01 (Christy et al. 2003), 0.1
(Mears et al. 2003), and 0.20 K/decade (Vinnikov
et al. 2006), we have a TTR trends of 0.09,
0.18, and 0.29 K/decade, respectively.
When is global warming really a cooling?
By Roy Spencer
Published 05/05/2004
http://www.techcentralstation.com/050504H.html
New climate study finds ‘global warming’ by
substracting cooling that wasn’t there
University of Alabama at Huntsville (UAH) News Release
05/05/2004
Assault from above
A Report Produced by The CO2 & Climate Team
Published 05/06/2004
http://www.co2andclimate.org/wca/2004/wca_17apf.html
Spencer (05/05/2004)
The Fu et al. weighting function shows substantial negative weight above
100 hPa, a pressure altitude above which strong cooling has been
observed by weather balloon data. This leads to a misinterpretation of
stratospheric cooling as tropospheric warming.
__ Spencer (05/05/2004)
Methodology
We use the observed vertical profile of stratospheric
temperature trend to directly evaluate the magnitude
of stratospheric contamination in various techniques
used to estimate the tropospheric temperature trends:
TÝ
200
 TÝ( p)W ( p)dp
0
Stratospheric Trend Profile
0
20
40
Pressure (hPa)
60
80
100
120
140
R_H
160
R_P
180
200
-1.2
HadRT
-1
#+
-0.8 -0.6 -0.4
Trend (K/decade)
ox
-0.2
0
Fig.1. Mean vertical profile of
temperature trend in the stratosphere as
compiled by Ramaswamy et al. (2001)
using radiosonde, satellite, and analyzed
data sets, rescaled to the global trend of
UAH MSU T4 over the 1979-2001
period. The solid and dashed lines
represent trend profiles using linear
extrapolation with respect to height and
pressure, respectively, below 15 km
(~120 hPa). Also shown are the global
temperature trends for the layer between
100 and 300 hPa for the same time span,
as derived from four radiosonde
datasets: Angell-63 (Angell-54 (+),
HadRT (o), and RIHMI (x) (See Seidel
et al. 2004 for detailed descriptions of
these datasets).
Stratospheric Contamination (K decade -1)
A Direct Error Estimates
0.02
0
-0.02
-0.04
R_H
-0.06
R_P
HadRT
-0.08
-0.1
W_2
W_2LT
W_FT
Fu and Johanson (2004, J. Climate)
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
According to the
published report, “there
is no longer a
discrepancy in the rate of
global average
temperature increase for
the surface compared
with higher levels in the
atmosphere. This
discrepancy had
previously been used to
challenge the validity of
climate models used to
detect and attribute the
causes of observed
climate change”.
Climate Change 2007: The Physical
Science Basis
Summary for Policymakers
Warming of the climate system is unequivocal, as is now
evident from observations of increases in global average air
and ocean temperatures, widespread melting of snow and
ice, and rising global average sea level (see Figure SPM-3).
…
New analyses of balloon-borne and satellite measurements of
lower- and mid-tropospheric temperature show warming rates
that are similar to those of the surface temperature record and
are consistent within their respective uncertainties, largely
reconciling a discrepancy noted in the TAR.
…
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
“One issue does remain
however, and that is related
to the rates of warming in
the tropics. Here, models
and theory predict an
amplification of surface
warming higher in the
atmosphere. However, this
greater warming aloft is not
evident in three of the five
observational data sets used
in the report. Whether this is
a result of uncertainties in
the observed data, flaws in
climate models, or a
combination of these is not
yet known.”
Water vapour feedback structure:major
contributions
BMRC model:
TOA Radiative impact of water
vapour changes, 2CO2 1CO2
(Wm-2K-1100hPa-1)
Courtesy of Colman
Some Basics on Tropical Tropospheric
Temperature Profiles
Uniform tropical upper-air
temperature
Larger SST
variations
Rainfall, cloud cover, and humidity all roughly follow warm
SST
Mechanisms Setting Tropical Temperature Distribution
In the tropical atmosphere:
• The troposphere is stably stratified outside convective regions.
• Release of latent heat in deep convective systems balances
radiative cooling and heat export.
• These systems keep the local temperature profile roughly
moist-adiabatic.
15 km
z
TLH
zLCL
T+gz/cp
• Moist static energy h = cpT + Lq + gz roughly conserved
in deep cumulus updrafts
• hsat = cpT+Lqsat(T,z)+ gz = hABL
• Cumulative latent heating TLH(z) = L{qsat(T)-qsat(TLCL)}/cp
For TLCL = 296 K, TLH = 40 K at z = 4.5 km (T=0).
• A change in Tsfc with constant relative humidity
changes the temperature profile like
dT/dTsfc = (1 + gLCL)/(1+g(T)) > 1, g = (L/cp)dqsat/dT
(lapse rate feedback). Currently gLCL = 3, g(273 K) = 1.2,
at the tropical freezing level, dTair/d(SST) = 1.9.
Courtesy of Bretherton
Stratified Adjustment
• Coriolis parameter f = 2W sin(latitude)
• Gravity waves efficiently spread heat over a Rossby
radius R = NH/f 
• This maintains a horizontally uniform temperature profile
over the entire tropics determined by moist adiabatic lifting
of near-surface air over warm moist parts of the tropics
(e.g., Charney 1963; Schneider 1977; Held and Hou 1980;
Bretherton and Smolarkiewicz 1989; Sobel and Bretherton 2000).
Q
z
C ~ 50 m s-1
T+gz/cp
T+gz/cp
Courtesy of Bretherton
ENSO Example: Warm-Phase SST
Anomalies
Vertical Structure of ENSO-Regressed Air
Temperature Variation IS nearly MoistAdiabatic
Enhanced upper
tropospheric warming
Chiang and Sobel (2002)
(Chiang and Sobel 2002)
Some Tropical Climate Basics
• In the deep tropics, air temperature is nearly
horizontally uniform above the atmospheric boundary
layer, which is coupled to warmest SSTs and roughly
moist-adiabatic vertically.
• The physics behind those seems suggest that they
probably also hold in changed tropical climates.
• Show supporting observations using current-day
climatology versus ENSO as an example ‘climate
variation’.
We might expect that across the tropics, tropospheric
temperatures would respond uniformly to climate
changes. They should be locked to warm tail of SSTs
and the T changes should amplify moist-adiabatically
with elevation.
Formulation of MSU Effective
Weighting Functions for Different
Tropical Tropospheric Layers
0
(a)
Pressure (hPa)
100
W3
W4
0
(b)
100
200
200
300
300
400
W2 (0.05)
500
600
W2LT (0.1)
400
WTT (0.055)
500
WTLT (0.08)
600
700
700
800
800
900
900
1000
1000
-0.003 0 0.003 0.006 0.009 0.012 -0.003 0 0.003 0.006 0.009 0.012
Weighting Function (1/hPa)
Weighting Function (1/hPa)
Fu & Johanson
(2005, GRL)
Temperature Trends (K/decade)
Vertical Structure of Tropical
Tropospheric Temperature Trends
0.5
30N-30S: 1987-2003
0.4
Ts: HadCRU2v
0.3
T TLT
0.2
Ts
TTT
TTT
Ts
0.1
0
T 2LT
-0.1
RSS
UAH
Fu and Johanson (2005, GRL)
TTLT-T2LT (K)
TTLT-T2LT (K)
Discussions on T2LT
0.55
(a)
0.45
0.35
0.25
0.15
0.05
-0.05
30N-30S: Ocean Only
-0.15
1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
0.55
(b)
0.45
0.35
0.25
0.15
0.05
-0.05
30N-30S: Land Only
-0.15
1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003
Year
Fu &Johanson
(2005, GRL)
UAH T2LT trend
bias is largely
attributed to the
periods when
satellites had
large local
equator crossing
time drifts.
Mears & Wentz (2005, Science)
SUMMARY NOTES
•
•
•
•
Trends in T2 are weak because the instrument
partly records stratospheric temperatures whose
large cooling trend offsets the contributions of
tropospheric warming.
We quantify the stratospheric contribution to T2
using MSU channel 4, which records only
stratospheric temperatures.
We find that the stratospheric contamination in T2
trend is -0.08 K/decade for the period from
1/1/1979 to 12/31/2001.
The results of Fu et al. (2004) are validated with
a direct error analysis and are also independently
repeated by Gillett, Santer & Weaver (2004,
Nature) and Kiehl et al. (2005).
•
•
•
•
The satellite-inferred tropospheric temperature
trends after removing the stratospheric
contamination are physically consistent with the
observed surface temperature trends.
The UAH T2LT trend in the tropics is physically
implausible, which is verified by Mears & Wentz.
We quantify the trend in tropical tropospheric
temperature vertical structure by using
combinations of MSU T2, T3, and T4.
The satellite-inferred tropical air temperature
trends based on RSS MSU data increase with
height.
Global Stratospheric & Tropospheric
Temperature Trends (1979-2005)
Qu ickTime™ an d a
TIFF (U ncom pre sse d) de com pres sor
are nee ded to s ee th is p icture .
Fu, Johanson, Wallace and Reichler (2006, Science)
Pole-ward Shift* of Tropospheric Jet
Streams from MSU Obs.
DJF
MAM
JJA
SON
NH
0.8
1.2
1.4
-0.2
SH
-1.6
-0.4
-0.8
0.0
Total
2.4
1.6
2.2
-0.2
*degree for last 27 years
Hadley Circulation Broadening Seen
from OLR
ERBS Edition 3_Rev1
Wong et al. (2006)
HIRS Pathfinder
Mehta & Susskind (1999)
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
AVHRR Pathfinder
Jacobowitz et al. (2003)
ISCCP FD
Zhang et al. (2004)
GEWEX RFA
Stackhouse et al. (2004)
Hadley Circulation Broadening Seen
from OLR since 1979
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
Hu and Fu (2007)
HIRS Pathfinder: 4.8o; ISCCP FD: 4.0o; GEWEX RFA: 2.3o
Hadley Circulation Broadening Seen
from Meridional mass stream function
Total expansion from
reanalyzes
ECMWF 2.6°
NCEP/NCAR 2.7°
NCEP/DOE 3.1°
Hartmann (Global Physical
Climatology, 1994)
Evolution of zonal mean
meridional mass stream
function at 500 hPa in the
northern hemisphere for
(SON) (Hu and Fu 2007)
Hadley Circulation Broadening
Seen Satellite observed ozone
TOMS total ozone field for March 11th,
1990.
Monthly mean relative areas
Tropical
Sub-arctic
Mid-latitude
Polar
Hudson et al. (2006)
Tropical and mid-latitude boundaries
separated by upper troposphere jet
Between 1979 and 2003, the
tropical regime expanded by
~2.7 degrees in the northern
hemisphere alone
Models versus Observations
Observed expansion (based on OLR) cannot be explained by natural variability
Expansion in GFDL model simulations is weak, non-existent, or in opposite direction as
observations
SUMMARY NOTES
•
•
•
•
Three reanalyses, three OLR datasets, satellite ozone
obs. and satellite MSU obs. in terms of MMS, OLR,
ozone, and tropospheric temperature trends all indicate a
significant broadening of Hadley circulation (~2 to 5o)
since 1979.
GCMs cannot reproduce the observed Hadley cell
expansions. The 21st century climate change simulations
of the IPCC AR4 suggest a robust pole-ward expansion of
the Hadley circulation (Lu et al. 2007) but they are much
weaker than those based on observations.
Important implication to midlatitude drought (e.g.,
Hoerling & Kumar 2003, Science; Lau et al. 2005).
An indication of GCMs’ inability to simulate Eocene
equator-to-pole surface temperature gradient???
Tropospheric Temperature Trends in
Antarctica (1979-2005)
•
•
•
•
Recent debates on the Antarctic climate change (Doran et
al. 2002, Nature; Turner et al. 2002, Nature; Jones &
Widman 2004, Nature; Bertler et al. 2004).
Antarctic cooling in the summer-fall season (Thompson &
Solomon 2002, Science; Shindell &Schmidt 2004).
Significant uniform Antarctic winter tropospheric warming
(Turner et al. 2006, Science).
No significant change in snowfall (Monaghan et al.
Science 2006), which seems inconsistent with winter
tropospheric warming.
Turner et al. (2006) used radiosonde data at nine stations over
Antarctic: “…satellite product (T2lt) may not be reliable around
Antarctica in the winter because of the effects of the sea ice.”
Comparison of Tropospheric
Temperature Trends between
Radiosonde and MSU (T2&T4)
Johanson & Fu (2007)
Trend Pattern in Antarctica
Troposphere
Stratosphere
Johanson and Fu (2006)
Summary Notes
• The tropospheric temperature trends retrieved from MSU T2
and T4 agree with those from eight Antarctic radiosonde
stations (but not at Bellingshausen where there is a large
false warming from the radiosonde).
• The Antarctic continent is cooling in summer-fall season
since 1979, which agrees with previous study.
• About half of the Antarctic continent is not warming but even
cooling in the winter, which does not support Turner et al.
(Science 2006) but is consistent with the snowfall change
reported by Monaghan et al. (2006, Science).
• We identify major stratospheric warming in part of the
Antarctica in the winter-spring season, which requires an
explanation.
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.