Earth Radiation Budget Studies

1
Earth Radiation Budget Studies

• Aaron B. Wilson
• May 28, 2009

2
Overview

• Why Earth Radiation Budget Studies Important?
• Reexamination of the Observed Decadal Variability
  of the Earth Radiation Budget Using
  Altitude-Corrected ERBE/ERBS Nonscanner WFOV Data
  Wong et al. 2006
• Toward Optimal Closure of the Earths
  Top-of-Atmosphere Radiation Budget Loeb et al.
  2009
• Common Summary and Conclusions

3
Earth Radiation Budget

• Weather and Climate dependent upon
• Amount of incoming radiation
• Distribution of incoming radiation
• Equilibrium-global net radiation at the
  top-of-the atmosphere (TOA) is 0
  absorbedemitted
• Various weather system and interactions work
  toward this balance
• Imbalance leads to adjustment within the climate
  system by warming or cooling the global
  temperature
• Small changes in the TOA fluxes lead to large
  changes within the climate system
• According to Hansen et al 2005 (Science)-imbalance
  of 1 Wm-2 maintained for the last 10,000 years
  of the Holocene melt ice equivalent of 1 km of
  sea level

4
Outline Manuscript 1

• Reexamination of the Observed Decadal Variability
  of the Earth Radiation Budget Using
  Altitude-Corrected ERBE/ERBS Nonscanner WFOV Data
  Wong et al. 2006
• Altitude change correction addressed
• Shortwave (SW) instrument drift addressed
• New data set compared against model
  Top-of-Atmosphere fluxes
• New data set compared to other satellite-based
  decadal TOA flux data sets
• ERBE and CERES/Terra FM1 compared with global
  ocean heat storage for a 10-year period
  (1993-2003)
• Results and conclusions drawn upon for
  suggestions for future ERB studies

5
Motivation

• Based on previous work on Earth Radiation Budget
  (ERB) by Wielicki et al. 2002
• Large decadal changes in tropical mean (20?N to
  20?S) earth radiation budget between 1980s and
  1990s
• Based on longest-running single ERB time series
  from Earth Radiation Budget Experiment
  (ERBE)/Earth Radiation Budget Satellite (ERBS)
  Nonscanner Wide Field of View (WFOV) instrument
• Also examined
• Nimbus-7 Nonscanner
• ERBE/ERBS Scanner
• Scanner for Radiation Budget (ScaRaB)-Meteor
• ScaRab-Resurs
• Clouds and the Earths Radiant Energy System
  (CERES) on TRMM
• CERES on Terra

6
Altitude-Corrected WFOV Ed 3
ERBS Satellite

• Nonscanner WFOV instrument
• Contains entire earth disk and ring of
  surrounding space
• Amount of energy received at instrument is
  inversely proportional to the square of the
  distance between the instrument and the earths
  center
• Altitude observed fluxes converted to TOA fluxes
  through inversion process
• Satellite dropped from 611km to 585km over 15
  year period
• Small increase in TOA fluxes 0.6

7
Altitude-Corrected WFOV Ed 3

• ERBE/ERBS Nonscanner WFOV Ed 2 reprocessed
• Time-dependent correction coefficients
• 36-day averaged tropical mean budget
• LW 3.1 ? 1.6 Wm-2
• SW -2.4 ? -3.0 Wm-2
• Net -0.7 ? 1.4 Wm-2

8
WFOV SW Instrument Drift

• ERBS WFOV instrument trend
• Due to non-uniform exposure of the WFOV SW dome
  to UV radiation during spacecraft sunrise and
  sunset
• Sides of SW sensor dome receive more exposure
  than top of dome
• Leads to difference in dome transmission
• No dedicated longwave sensor Day LW Tot SW
  so both SW/LW affected

• SW dome has degraded allowing less solar energy
  onto the detector producing a lower SW flux and
  higher LW flux during the daytime
• LW 1.6 ? 0.7 Wm-2
• SW -3.0 ? -2.1 Wm-2

9
MFOV SW Instrument Drift

• ERBS MFOV instrument
• 800km diameter FOV
• Adjacent to the WFOV
• Shielded from the sun
• Therefore no trend
• Noise much larger covers tropics only every 4
  days
• Help support the non-uniform exposure theory of
  the SW dome

10
WFOV Uncertainty

• Calibration Stability
• Stability in solar calibrations of 0.1 0.35
  Wm-2
• Annual mean spatial sampling error lt 0.1 Wm-2
• ERBS requires 2 days to view earth 60?S to 60?N
• Angular sampling error 0.2 Wm-2
• Time sampling error SW 0.1 Wm-2/LW 0.1 Wm-2

11
New ERBS vs. Models

• Models included
• Hadley Center Atmospheric Climate Model version 3
  (HadAM3)
• NCAR Community Climate Model version 3 (CCM3)
• Geophysical Fluid Dynamics Laboratory (GFDL)
• GFDL Experimental Prediction (EP)
• NCEP-NCAR 50yr Reanalysis
• Good agreement in the LW
• SW and Net still show differences
• Note El Nino -1998 and Mt Pinatubo in 1991-1993

12
New ERBS vs. Other Satellite ERBs

• Satellite-based decadal data used
• HIRS-High resolution Infrared Radiation Sounder
  Pathfinder LW
• International Satellite Cloud Climatology Project
• Advanced Very High Resolution Radiometer
  Pathfinder ERB dataset
• AVHRR consistently different intercalibration
  and satellite orbit changes throughout the period

13
ERBS/CERES vs. Ocean Heat Storage
ERBS (60-60) and CERES (global) 12 month running
means compared to 85-89

• Global ocean heat storage for 1992-2002
• Improved in situ temperature profile sampling and
  global altimeter data
• Change in TOA radiation should be same magnitude
  and in phase with ocean heat storage
• Ocean 10x more heat
• In phase and show interannual variability
  correlation

Willis et al. 2004
14
Overlapping Climate Record

• Tropical means LW anomaly w/r/t 1985-89
  climatology
• Scanner and nonscanner disagree but w/i absolute
  accuracy of instruments
• Demonstrates the need in overlapping climate data

15
Summary and Suggestions

• After altitude correction and sensor dome
  degradation correction, new ERBS Nonscanner WFOV
  dataset agrees well with models and existing
  decadal ERB datasets
• Agreement with Ocean Heat Storage
• Ocean heat storage and net radiation interannual
  variability are consistent with heating predicted
  from coupled ocean-atmosphere climate models (not
  shown)
• Variability will require long and accurate time
  series of heat storage and clear-sky, all-sky,
  and could radiative forcing observations for
  cloud feedback studies
• Need advanced instrument calibration, reduced
  gaps in climate data, and independent ERB
  observations with independent analysis

16
Outline Manuscript 2

• Toward Optimal Closure of the Earths
  Top-of-Atmosphere Radiation Budget Loeb et al.
  2009
• Description of several data sets used for TOA
  flux comparisons including 3 CERES data sets
• Analysis of various uncertainties involved in
  satellite-based radiation studies
• CERES data set SW and LW adjusted to remove
  inconsistency between fluxes and heat storage
• Description of combined CERES-MODIS data set for
  high-resolution clear-sky fluxes
• Regional comparisons between adjusted ERBE and
  CERES radiative fluxes
• Summary and Conclusions including conclusions
  drawn between both papers

17
Introduction

• The rate at which the Earth reacts to an
  imbalance energy is modulated by its capacity to
  store energy
• Previous paper showed the phase and magnitude
  likeness to TOA and Ocean heat storage
• Models-imbalance has grown since 1960s and Earth
  now absorbs 0.85 0.15 Wm-2 more energy than it
  releases back to space
• Based on Hansen et al. 2005 Used Goddard
  Institute for Space Science with GHG forcing
• Köhl and Stammer 2008 40 more due to deep ocean
  heat content change

18
A Closer Look at 0.85 Wm-2

• 5 run mean using GHG forcing, solar irradiance,
  albedo, aerosols, and land use to determine TOA
  net radiation

Hansen et al. 2005-Science
19
Introduction

• Fasullo and Trenberth (2008) show much larger net
  imbalance from satellite measurements
• Show more stability over time than absolute
  calibration accuracy
• Must account for difference in long term net
  radiation and heat storage on Earth
• Early attempts Larger adjustments made to SW
  than LW-sampling and modeling of diurnal cycle

20
Data

• Data sets include gridded monthly mean TOA
• ERBE broadband radiometer
• CERES-Terra broadband radiometer
• Filtered radiances in SW, Total, and Window
  wavelengths
• Converted to unfiltered and LW daytime Total -
  SW
• Global Energy and Water Cycle Experiment (GEWEX)
  Surface Radiation Budget (SRB)
• International Satellite Cloud Climatology Project
  (ISCCP)radiative transfer model calculations
• 5yr period (March 2000-February 2005)-except ERBE
  (February 1986-January 1989)
• Simultaneously on ERBS and NOAA-9 or-10

21
More About CERES Data

• 3 CERES sets
• ERBE like (same algorithms) ES-4
• SRBAVG-NonGeo monthly TOA/surface averages
  nongeostationary
• Spatially average instantaneous values in
  equal-area, temporally at 1-h increments for
  entire month, and average all hour boxes in the
  month
• SRBAVG-GEO
• Both SRBAVG 1? x 1? with MODIS cloud and aerosol
  properties
• 3hourly visible-infrared from 5 Geo satellites to
  account for observations between CERES
  measurements
• CERES Single Scanner Footprint TOA/Surface Fluxes
  and Clouds (SSF)
• Merges CERES with MODIS characterized clear and
  cloudy portions of CERES footprint

22
Global/Regional Mean TOA Fluxes

• LW Range 4.6 Wm-2 SW Range 8.6 Wm-2
• CERES SW about 3 Wm-2 lower than rest
• All-Sky Net TOA Range 7.3 Wm-2

23
CERES Regional Differences

• A,C,E Show differences between ERBE-like and
  SRBAVG-NonGeo
• Difference due to angular distribution models
  used to convert unfiltered radiances to TOA
  fluxes
• B,D,F Show difference between NonGeo and Geo
• Difference due to temporal interpolation

24
Sources of Uncertainty SRBAVG-GEO

• According to the Total Irradiance Monitor on
  SOURCE Satellite irradiance 1361 0.8 Wm-2
• 1365 Wm-2 used in this study (based prior to
  SOURCE) 4 Wm-2 globally average 1 Wm-2 bias
• Assume spherical earth (instead of oblate
  spheroid) 0.29 Wm-2 bias net TOA flux
• Terminator Error bias 0.3 Wm-2 extrapolating
  albedo at solar zenith angles from 75-85? to
  85-90? after instantaneous value at 90? on
  CERES-TRMM
• Absolute Calibration 2 SW and 1 TOT ? 4.2
  out of 6.5 Wm-2 in net due to this bias

25
SRBAVG-Geo TOA Flux Uncertainties
26
Adjusting CERES TOA Fluxes

• Constrainment algorithm used to identify these 12
  error sources for TOA fluxes
• Remember Assumed 0.85 Wm-2 imbalance exits
• 90 of error is absolute calibration of instrument

27
Adjusted CERES TOA Fluxes

• ERBE and CERES adjusted show differences

28
ERBE(08) CERES TOA Fluxes

• SW maximum difference in higher latitudes
• LW ERBE generally lower than CERES
• Net shows High Latitudes
• Failure to correctly identify clear-sky snow and
  sea ice from clouds in ERBE

SW
LW
NET
Solar Irradiance
29
High-resolution Clear Sky Fluxes

• CERES clear-sky at 20km footprint missing
  smaller scale radiative cloud effects
• Area weighted average
• CERES cloud free
• MODIS derived clear sky from partly/mostly cloudy
  footprints of CERES-MODIS narrowband radiances
  converted to broadband SW fluxes
• March 2002 monthly mean results
• SW and LW bias from conversion
• MODIS-CERES fluxes
• MODIS-CERES and CERES-only differences
• SW 0.9 Wm-2
• LW - 0.3 Wm-2

30
ERBE(97) -CERES ASR TOA/Zonal

• All-Sky ERBE is lower than CERES
• Clear-Sky over sea ice and snow an issue in ERBE
• CERES uses MODIS and SSM/I

31
ERBE(97) -CERES LW TOA/Zonal

• LW fluxes smaller in ERBE than CERES

32
ERBE(97)-CERES Net TOA/Zonal

• Generally less Net in ERBE especially in the
  tropics
• Large difference over Southern Hemisphere

33
Summary and Conclusions

• Best estimate of imbalance is 0.85 Wm-2 (GISS
  model)
• Satellites show larger imbalances due mainly to
  absolute calibration, but other factors involved
• Solar constant, spherical earth, radiance to flux
• Small scale radiative cloud effects show subtle
  changes through improved MODIS-CERES application
• After adjusted for error, CERES and ERBE adjusted
  show large differences in TOA fluxes

34
Drawing It All Together

• We have seen there is a large amount of
  uncertainty to consider when examining the ERB
• Satellite issues Altitude changes, sensor
  degradation, instrument calibration,
  intercalibration between subsequent records
• global net imbalance assumptions
• Improved instrument calibration and overlapping
  climate records should help improve results
• The signal being sought is extremely small, and
  every source of error must be considered