Microwave Remote Sensing: 2' Theory, Algorithms and Applications

1
Microwave Remote Sensing2. Theory, Algorithms
and Applications

• Dr. Fuzhong Weng
• Sensor Physics Branch
• Center for Satellite Applications and Research
• National Environmental Satellites, Data and
  Information Service
• National Oceanic and Atmospheric Administration
• 2009 Update

2
Outline

• Microwave Remote Sensing Theory
• MW gas spectrum
• Radiative Transfer Approximation
• Microwave Algorithms
• Cloud liquid water
• Cloud ice water
• Precipitation
• Temperature and water vapor
• Product Applications
• Model validation and intercomparison
• NWP model validations
• Climate monitoring and trending

3
Microwave Absorption Spectrum
1. Rotational transition line O3,H2O,CO,ClO,
N2O 2. Spin-rotational transition O2 and
zeeman splitting in upper atmosphere where
geomagnetic field is important 3. Doppler and
pressure broading
Janssen M. A., 1993 Atmospheric remote sensing
by microwave radiometry, Chapter 2, John Wiley
Son inc
4
Microwave Penetration Depth
5
Instrument Spectrum Allocations
6
MW Stratosphere and Mesosphere Sounding
7
Millimeter Wavelength Spectroscopy
8
Global Land Emissivity Characterization SSM/I
Fifteen Year Time Series

• Large season change at higher frequencies
• Large polarization difference for several
  surfaces (e.g. desert, snow, flooding)
• Deserts appear as a scattering medium

SSM/I surface emissivity climatological data set
is developed at various time scales (e.g. pentad,
weekly and monthly, anomaly). SSM/I sensors from
F10 to 15 satellites are intercalibrated to a
reference satellite (F13)
9
Microwave Surface Emissivity Spectra
10
Advanced Microwave Sounding UnitImaging and
Temperature Sounding Channels
31.4 GHz
23.8 GHz
53.7 GHz
52.8 GHz
11
Advanced Microwave Sounding UnitImaging and
Moisture Sounding Channels
1833 GHz
89 GHz
150 GHz
183 1 GHz
12
Outline

• Microwave Remote Sensing Theory
• MW gas spectrum
• Radiative Transfer Approximation
• Microwave Algorithms
• Cloud liquid water
• Cloud ice water
• Precipitation
• Temperature and water vapor
• Product Applications
• Model validation and intercomparison
• NWP model validations
• Climate monitoring and trending

13
Microwave Remote Sensing of Clouds

• A large contrast exists between cloudy and
  clear conditions, thanks to low ocean
  emissivity.
• Brightness temp increases exponentially with
  liquid water, thus requiring a logarithmic
  function for linearization
• The linear regime is dependent on frequency.
  We can meet more customers needs (e.g. rain
  water..) if the measurements at each frequency
  are optimally utilized in the retrievals

14
Emission Approach
15
Emission-Based RT Model (1/3)
16
Emission-Based RT Model (2/3)
17
Emission-Based RT Model (3/3)
18
Why MW cant see cloud well over land?
19
Liquid Water Absorption
20
Scattering Approach 2 Streams Approximation
21
Two-Stream Model Solution
22
Algorithms of Cloud (Rain) Liquid Water Path
Vertically Integrated Liquid Water over Unit Area
23
Cloud Liquid Water Algorithm
Sometime, satellite measurements under clear
condition can be used to derive
some coefficients. From Eq. 6.13, set L0
24
Cloud Liquid Water Algorithm Evolution
Statistical all coefficients are from simulated
data sets Ts290
Emissision-Based
SemiPhysical a0 is derived from simulated data
set, a1 and a2 are from microwave measurements
under clear conditions
a0 -0.5?v23 / (?v23?l31 - ?v31?l23) b0
0.5?l23 / (?v23?l31 - ?v31?l23) a1 ?v31 /
?v23 b1 ?l31 / ?l23 a2 - 2.0(?o31 - a1
?o23) / ? (1.0 - a1) ln(Ts) ln(1.0 -
?31 ) a1 ln(1.0 - ?23 ) b2 -
2.0(?o31 - b1 ?o23) / ? (1.0 -
b1)ln(Ts) ln(1.0 - ?31 ) b1 ln(1.0
-?23)
Physical all coefficients are derived as
function of cloud layer temp, surface wind speed,
surface temp
Scattering Based
Physical Microwave integrated Retrieval system,
1dvarwith scattering RT, all hydrometeors
profiles
25
SSM/I Cloud Liquid Water Algorithm Operational
at FNMOC and NESDIS

• Pros
• Semi-Physical with easy understanding
• Large dynamic range (rain and non-rain)
• Clean background due to uses of real measurements
• Validated with ASTEX data for non-raining clouds

• Cons
• Difficult to accommodate information from
• new channels and ancillary data
• Cloud layer temp is implicit

26
(No Transcript)
27
NOAA POES AMSU

• AMSU are on board NOAA POES since 1998
• There are 20 channels divided into three
  sub-modules
• A1 13 channels located near the 60 GHZ oxygen
  absorption band
• A2 2 window channels at 23.8 and 31.4 GHz
• B 2 high frequency channels at 89 and 150 GHz,
  and 3 channels near 183 GHz water vapor
  absorption line
• The field-of-view size varies as the instruments
  scan crossing track

28
AMSU Weighting Functions
29
NOAA-16 AMSU-A Radiance Asymmetry (Channel
1,2,3,15)
,
?T A0 exp -0.5(? - A1) /A22 A3 A4 ?
A5 ?2
30
NOAA-15 AMSU-A Radiance Asymmetry (Channel
1,2,3,15)
,
?T A0 exp -0.5(? - A1) /A22 A3 A4 ?
A5 ?2
31
Cloud Absorption in relation to Temperature
32
AMSU Cloud Liquid Water
Before Asymmetry Correction
After Asymmetry Correction
NOAA-15
NOAA-16
33
Algorithms of Cloud Ice Water Path Vertically
Integrated Ice Water over Unit Area
34
Cloud Ice Water Path Algorithm
I0
?0
B(T)
?1
I1
Weng and Grody (2000, JAS) Zhao and Weng (2002,
JAM)
Asymptotic Limits
1. Emission Approach
2. Scattering Approach
35
Definitions of Cloud Ice Water Path
36
CIWP Error Budget
CIWP retrieval errors ()

• The errors of CIWP are mainly due to
• uncertainty in the effective particle diameters
• uncertainty in the particle bulk volume density

?De/De()
CIWP retrieval errors ()
??i/?i ()
37
ER-2 MIR, DC-8 ARMAR, MODIS Simulator
Measurements
Three millimeter wavelength channels provide the
overall needed sensitivity for cloud ice
microphysics which can be uniquely used for
precipitation mapping
Weng and Grody (2000, JAS)
38
MIR Window Sounding Channel Observations
39
Flowchart of Cloud Ice Algorithm
40
Cloud Ice Water Path

• Brightness temperatures from AMSU-B 89 and 150
  GHz are two primary channels for IWP and De
• Retrieval algorithm was published in Journal of
  Atmos Sci (Weng and Gody, 2000) and J. Appli.
  Meteor (Zhao and Weng, 2002)
• AMSU-A window channels are used for surface
  screening.
• The algorithm works for opaque ice clouds having
  IWP greater than 0.05 kg/m2

41
Algorithms of Atmospheric Sounding
42
MIRS System Design Architecture
External Data Tools
Raw Measurements Level 1B Tbs
Radiance Processing
Radiometric Bias
Ready-To-Invert Radiances
Inversion Process
NEDT Matrx E
RTM Uncert. Matrx F
Geophysical Bias
NWP Ext. Data
EDRs
In-Situ Data
Comparison
43
Cost Function Minimization

• To find the optimal solution, solve for
• Assuming linearity
• This leads to two iterative solution
• when number of parameters ltlt channel
• Otherwise

44
Microwave TPW Extended over Land
GDAS Analysis
MIRS Retrieval
45
Validation of TPW Retrieval over Land

• 4000 NCDC IGRA points collocated with NOAA-18
  radiances
• Only convergent points over land used
• Only points within 0.5 degrees and within 1hour
• Cloudy points included, up to 0.15 mm

46
Global Temperature Profiling
No Scan-Dependence in retrieval Smooth Transition
Land/Ocean
QC-failure is based on convergence Focus of
on-going work
47
Global Humidity Profiling
No Scan-dependence noticed Angle dependence
properly accounted for
48
Outline

• Microwave Remote Sensing Theory
• MW gas spectrum
• Radiative Transfer Approximation
• Microwave Algorithms
• Cloud liquid water
• Cloud ice water
• Precipitation
• Temperature and water vapor
• Product Applications
• Model validation and intercomparison
• NWP model validations
• Climate monitoring and trending

49
Microwave Environmental Data Records
50
Intercomparison between MODIS and AMSR-E LWP for
Stratus Clouds
51
Validation of General Circulation Model
52
Validation of Numerical Weather Prediction Models
53
(No Transcript)
54
GFS Prognostic Scheme vs. AMSU Cloud Water
Satellite Retrievals
GFS model (non-raining)
Eta model (non-raining)
Eta model (rainingnon-raining)
It is obvious that global/regional models have
ice happy physics
55
Impacts of SSMIS LAS on Hurricane Temperature
Analysis
Test
Control
Liu and Weng, GRL, 2007
56
Katrina Warm Core Evolution
57
Cloud Ice and Precipitation Distribution
Hurricane Dean (2007)

• SSMIS Rain Rate Retrieval Algorithm Summary
• Derive IWP and De from SSMIS 91 an 150 GHz using
  two-stream model (Weng and Grody, 2000)
• IWP is then converted to surface rain rate (Weng
  et al., 2002)
• SSMIS derived surface snow and sea ice is used to
  screen the false signature

58
Typhoon Luosha
59
2008 IOWA Flooding
60
Summary

• Sounding and Imaging
• Profiling atmosphere and imaging clouds, precip
  and surface
• Microwave Absorption Bands
• O2 (50-60 Ghz, 117-120 GHz),
• H2O (176-190 GHz)
• Channels near the line centers for sounding
• Channels between the lines for imaging
• Microwave Cloud Algorithms
• Emission cloud liquid water/total precipitable
  water
• Scattering ice water path/ particle size
• Microwave Sounding Algorithms
• simultaneous retrievals from 1dvar
• All weather profiling requires scattering rt
  model
• Accurate surface emissivity model
• Main Applications
• NWP data assimilations
• Hurricane monitoring
• Surface flooding