AT737

1
AT737

• Precipitation I
• Stan Kidder

2
Measurement Approaches

• Visible and infrared techniques
• Only the outside of the clouds are sensed
• High resolution data are available
• Passive microwave techniques
• Hydrometeors are directly sensed
• but so are a lot of competing parameters
• Only low resolution data are available
• Radar
• Great sensitivity to precipitation
• Poor resolution
• and a bit on severe thunderstorms

3
VIS IR Techniques

• Cloud indexing techniques
• Different kinds of clouds rain at different rates
• Multispectral techniques
• Different rain rates have different spectral
  signatures
• Life-history techniques
• At different times in their life cycles, clouds
  rain at different rates
• Cloud model techniques
• Cloud models can be used to calibrate satellite
  observations

4
Cloud Indexing
Assign a rain rate to each cloud type
After Follansbee (1973)
R total rainfall ri rain rate for cloud type
i fi fraction of area or time covered with
cloud type i
Sum the rainfall
5
Cloud Indexing
GOES Precipitation Index (Arkin 1972)
GPI r f Dt
GPI total rainfall in an area r 3 mm/hr f
fraction of area with IR TB period (hours) for which f applies
6
GOES Precipitation Index
Mean Annual Precipitation (m) 1986-1995
Source http//tao.atmos.washington.edu/data_sets/
gpi/
7
Multispectral Techniques
Dittberner and Vonder Haar (1973)
P c1E c2A Po
P percent of normal seasonal precipitation E
seasonal mean infrared radiant exitance (OLR) A
seasonal mean visible albedo
Brighter and colder clouds rain more than others
8
GOES Multispectral Rainfall Algorithm
Uses 5 GOES Imager channels 0.65, 3.9, 6.7,
10.7, 12 mm
http//orbit-net.nesdis.noaa.gov/arad/ht/ff/gmsra.
html
9
GOES 10.7 mm
10
Radar
11
Life-History Techniques
Not used much due to computational complexity
12
Cloud Model Techniques
Convective-Stratiform Technique (Adler Negri
1988)

• Run a 1D cloud model (with a suitable sounding)
  to get a relationship between cloud-top
  temperatures and rain rate/area.
• Find the local minima in the IR image, which are
  assumed to be overshooting tops.
• Assign convective rain to each minimum as per the
  cloud model
• Assign stratiform rain to the anvil top
  (surrounding the IR minima).

13
Cloud Model Techniques
NESDIS Auto-Estimator
Started from a manual technique Scofield
Oliver (1977)
http//orbit-net.nesdis.noaa.gov/arad/ht/ff/auto.h
tml Based on Vicente et al. (1998). See
http//orbit-net.nesdis.noaa.gov/arad/ht/ff/techni
que.html
14
Passive Microwave Techniques

• Absorption schemes
• Scattering schemes
• Bayesian/RT modeling schemes

15
The Microwave Spectrum
Precipitation is measured in atmospheric windows
16
Microwave Basics

• Radiance is proportional to temperature, so
  Brightness Temp. (TB) is used instead of radiance
• Water surfaces have a low emissivity (0.5), so
  water surfaces have low TB (150 K)
• Land has a high emissivity (0.95), so land is
  warm (bright) in microwave images

17
89 GHz
Over ocean, clouds and water vapor brighten the
signal, rain scatters and lowers the signal.
18
150 GHz
Scattering by rain is more pronounced at 150 GHz
19
TB-Rain Rate Relationships

• Over the ocean, TB rises with rain rate due to
  absorption, then falls due to scattering.
• Over land TB falls with increasing RR.
• Higher frequencies are more sensitive to
  scattering and have a lower RR saturation point.

20
AMSU-B Rain Rate
Calculated with a scattering algorithm
21
A Bayesian Scheme

• Calculate TBs for a bunch of different rain
  profiles
• Using an observed TB vector (several
  frequencies), look up the closest rain profile
  (or average the closest rain profiles) to yield a
  rain profile retrieval (rain rate vs. height).

This is the way that rain retrievals are done on
TMI
22
Radar
Radar is the best! Why wait for Mother Nature to
send you a signal?
but
23
The View from the Top
The TRMM Precipitation Radar is the only one so
far
is not optimal
Horizontal scanning is difficult, which limits
the swath width.
24
Severe Thunderstorms

• Have rapidly growing anvils
• Have very cold overshooting cloud tops
• Grow vertically very fast

25
Anvil Growth Rate
26
Coming Attractions

• Look for the latest in satellite precipitation
  estimation in Chris Kummerows lecture