Return to previous page

NOTE: Please use the "Page Setup" item from the "File" menu in your browser to set the orientation to "Landscape" format (11" wide by 8.5" high) to insure the document prints correctly.

 

1.0  Overview

1.1  The Hydrologic Cycle
1.2  Advantages of Satellite Remote Sensing
1.3  Microwave Products
1.4  Microwave Spectrum
1.5  Channel Selection
1.6  Atmospheric Applications
1.7  Summary

2.0  Atmospheric Microwave Products

2.1  Microwave Water Vapor Imagery
2.2  Total Precipitable Water Definition
2.3  Total Precipitable Water Example
2.4  Water Vapor at Different Heights
2.5  GOES vs. Microwave Depiction of Moisture
2.6  Total Precipitable Water Question
2.7  Total Precipitable Water Feedback
2.8  Cloud Liquid Water Definition
2.9  Cloud Liquid Water Example
2.10   Cloud Liquid Water Question 1
2.11  Cloud Liquid Water Feedback 1
2.12  Cloud Liquid Water Question 2
2.13  Cloud Liquid Water Feedback 2
2.14  Summary

3.0  Precipitation Estimation Using Microwave Data

3.1  Introduction
3.2  Microwave and Infrared Imagery Comparision
3.3  Microwave Rain Rate Example
3.3a Flyout: Other Rain Measuring Techniques
3.4  Spectral Selection for Precipitation Estimation
3.5  Microwave Frequencies for Precipitation Estimation
3.6  Model of a Typical Convective Rain Cloud
3.7  Infrared Remote Sensing of Clouds and Precipitation
3.8  Microwave Remote Sensing of Clouds and Precipitation: Emission Frequencies
3.9  Microwave Remote Sensing of Clouds and Precipitation: Scattering Frequencies
3.10  Active Remote Sensing for Precipitation Estimation
3.11  Comparing Ground and Satellite-Based Radar with Passive Microwave Data
3.12 Summary

4.0  Case Example: Hurricane Ivan

4.1  Hurricane Ivan in the Atlantic
4.2  Comparing Dropsonde and SSM/I Data from Hurricane Ivan
4.3  Dry Air Interaction with Hurricane Ivan
4.4  Summary

5.0  Case Example: Southern California Flooding, January 2005

5.1  Introduction
5.2  500-hPa Heights over the West Coast and Pacific Ocean
5.3  GOES IR Window over the West Coast and Pacific Ocean
5.4  GOES Water Vapor over the West Coast and Pacific Ocean
5.5  AMSU TPW over the West Coast and Pacific Ocean
5.6  SSM/I TPW over the West Coast and Pacific Ocean
5.7  SSM/I Rain Rates over the West Coast and Pacific Ocean - I
5.8  SSM/I Rain Rates over the West Coast and Pacific Ocean - II
5.9  Radar Rain Rates over Southern California
5.10  GOES Hydroestimator Rain Rates
5.11  Precipitation Validation
5.12 Summary

6.0  Case Example: Heavy Precipitation in Alaska

6.1  Introduction
6.2  Large-Scale Synoptic Weather Pattern
6.3  GOES Water Vapor and AMSU Precipitable Water
6.4  Numerical Model Depiction and Snowfall Forecast
6.5  AMSU TPW and Anchorage Radar: Snowfall Begins
6.6  Summary

7.0  Case Example: Hurricane Katrina Rain Rates

7.1  Introduction
7.2  Passive Microwave Rain Rates of Hurricane Katrina
7.3  Doppler Radar Rain Rates of Hurricane Katrina over Florida
7.4  SSMIS Microwave Rain Rates of Hurricane Katrina over Florida
7.5  TMI Rain Rates of Hurricane Katrina over the Caribbean
7.6  Doppler Radar Rain Rates of Hurricane Katrina over New Orleans
7.7  Combining Polar and Geostationary Data for Optimal Rain Rate Products
7.8  Experimental Rain Rate Product Examples
7.9  TRaP Precipitation Estimation Technique
7.10  TRaP Example
7.11  Summary

8.0  Precipitation Monitoring Missions

8.1  The Tropical Rainfall Measuring Mission
8.2  TRMM Instruments and Scan Patterns
8.3  TRMM Detects a Squall Line
8.4  TRMM Tropical Cyclone Example
8.5  Global Precipitation Monitoring
8.6  GPM Dual-Frequency Radar Channels
8.6a  Flyout: Passive Microwave Remote Sensing of Polar Regions
8.7  GPM GMI Features
8.7a  Flyout: More on SSMIS Channels
8.8  GPM Applications: Rain vs. Snow
8.9  GPM Applications: Precipitation Examples
8.10  Summary

9.0  Summary

10.0  References

Return to Top


1.0  Overview

This module provides an introduction to microwave products that depict moisture in the atmosphere and precipitation rates. It begins with an explanation of total precipitable water and cloud liquid water and contrasts them with infrared water vapor imagery. Then a series of case examples are presented, highlighting the contributions TPW and microwave precipitation rate imagery make to accurate forecasting of weather systems. Finally, an introduction to the Global Precipitation Monitoring mission is presented. This module takes about one hour to complete.

Objectives:

After completing this module learners will be able to:

Table of Microwave Sensor Acronyms

Image that depicts a list of Polar-orbiting Environmental Satellite Microwave Instruments

Return to Top


1.1  The Hydrologic Cycle

thunderstorm photo over the US Midwest

"Man—despite his artistic pretensions, his sophistication, and his many accomplishments—owes his existence to a six inch layer of topsoil and the fact that it rains." - Anonymous

Animation of the hydrologic cycle

Click to view animation

This module provides an overview of products created with microwave remote sensing data that characterize the hydrologic cycle. Solar energy initiates the hydrologic cycle, evaporating surface water into water vapor—the first key atmospheric component that we will study. Then, lifting produces clouds—the second key component. The third and final atmospheric component is precipitation, which returns the water to the surface. Because they penetrate the atmosphere more effectively than visible and infrared sensors, microwave sensors on polar-orbiting satellites provide unique advantages for observing all facets of the hydrologic cycle.

Return to Top


1.2  Advantages of Satellite Remote Sensing

NPOESS 3 orbit animation placeholder
Click to view animation

Satellite observation of the hydrologic cycle can provide vital data over the world's otherwise largely unobserved oceans. Even over many relatively uninhabited land areas, satellites estimates are crucial for forecasting and global climate studies.

Return to Top


1.3  Microwave Products

AMSU global product composite

While microwave products depict both atmospheric and surface features, this module will focus on atmospheric products. Land and sea surface products will be covered in another module. The most important microwave products for atmospheric analysis include total precipitable water, cloud liquid water, and rain rates. Brightness temperature imagery at different frequencies is also used for analysis and forecasting.

Return to Top


1.4  The Microwave Spectrum

Electromagnetic Spectrum

 

When operational weather satellites were first launched, sensors relied mainly on visible and infrared sensing channels at relatively short wavelengths. Microwave remote sensing utilizes much longer wavelengths, expressed in units of frequency called gigaHertz (GHz). Meteorological observation frequencies fall in the range from about 1 to 300 GHz (30 to 0.1 cm). This range allows us to observe information about clouds and precipitation, monitor land and sea surfaces, and profile atmospheric temperature and humidity.

Return to Top


1.5  Channel Selection

microwave portion of the electromagnetic spectrum

Microwave channels are chosen at frequencies with different sensing capabilities. Some channels are located in window regions where there is little absorption by atmospheric gases. These channels help us identify properties associated with earth and ocean surfaces as well as clouds. Other channels, located in absorption regions, are sensitive to a range of atmospheric gases like water vapor, carbon dioxide, and oxygen. Channels in both window and absorption regions can be used to derive a variety of products including cloud properties.

Return to Top


1.6  Atmospheric Applications

AMSU microwave products composite

We will concentrate on three atmospheric elements of interest to weather forecasting: cloud microphysics, precipitation, and atmospheric water vapor. First we will cover some fundamental principles relevant to microwave remote sensing as it applies to the atmospheric products covered. Then we will consider several case examples using microwave products. We will also provide an overview of the Global Precipitation Measurement mission, reviewing current and future sensors that contribute to this effort.

Return to Top


1.7  Summary

Return to Top


2.0  Atmospheric Microwave Products

2.1   Microwave Water Vapor Imagery

AMSR-E 23 GHz Vertical Polarization image of the West Coast of the US

We will start by examining water vapor off the west coast of the United States. Using the 23-GHz channel from the Advanced Microwave Scanning Radiometer, or AMSR-E, we see a water vapor plume off California. The diffuse appearance of the plume confirms that we are looking at water vapor and not clouds. Notice that the plume disappears over land. This illustrates a general principle in satellite microwave remote sensing: detection and measurement of atmospheric phenomena is far easier over the ocean than over land.

microwave portion of the electromagnetic spectrum

This frequency is within a water vapor absorption region and provides data to create microwave water vapor products.

Return to Top


2.2  Total Precipitable Water Definition

tpw conceptual animation

tpw conceptual animation
Click to view animation

Because microwave instruments are able to sense energy through most clouds, we can create products that depict atmospheric constituents from the sea surface to the top of the troposphere. Total precipitable water (TPW) is a microwave product that represents the depth of liquid water that would be accumulated if all the water vapor in a hypothetical cylinder above a location on the earth were condensed into an equivalent amount of liquid water. Traditional infrared water vapor imagery using the 6.7-micrometer channel detects water vapor mainly at middle and high levels of the troposphere. TPW imagery on the other hand depicts the moisture present at all levels, especially important for the low levels where most water vapor is concentrated.

Return to Top


2.3  Total Precipitable Water Example

AMSU TPW global composite

TPW is one of the most accurate and useful passive microwave products. The Advanced Microwave Sounding Unit (AMSU), flown on the current generation of NOAA polar-orbiting satellites, provides coverage of a particular location about every six hours. Valid over oceans only, it is generally considered to be as accurate as rawinsonde values of integrated vapor. Expressed in millimeters, TPW has large values near the equator, where the sea is warm and evaporation rates high (shown here in red) and low values near the poles where there is little evaporation (shown here in greens and grays). Gradients in TPW can show the position of atmospheric fronts over the oceans.

Return to Top


2.4  Water Vapor at Different Heights

conceptual drawing showing high and low level water vap;or moving in different directions

Infrared water vapor, as seen on geostationary satellite loops, is an excellent tracer of motion at high and mid-levels in the atmosphere. In contrast, the water vapor depicted in TPW products is concentrated in the lower layers and may be moving in a different direction from moisture at higher levels. Forecasters may misdiagnose atmospheric conditions if they try to use conventional infrared water vapor imagery to follow low-level moisture plumes. As we will see, microwave products are well suited for tracking low- and mid-level moisture plumes.

Return to Top


2.5  GOES vs. Microwave Depiction of Moisture

SSM/I image of The Carribbean Basin

Let's compare polar orbiting microwave data with geostationary imagery. TPW is calibrated to sense the total moisture from the bottom to the top of the troposphere and has a spatial resolution of about 21 to 45 kilometers. Sometimes TPW products are given in inches, but more often in millimeters. This example provides both units. Valid everywhere over the oceans except in the thickest rain clouds, the product indicates the location of the highest low- and mid-level tropospheric humidity, the moisture responsible for heavy precipitation. Depending on the number of satellites in orbit, a new pass appears every 3 to 12 hours. In this instance the TPW product captures a major moisture plume moving into Louisiana, supplying moisture for flooding rains along the Gulf Coast.

GOES water vapor image of The Carribbean Basin

The corresponding GOES water vapor image is not very helpful in this instance. First, it cannot show us the flood-producing water vapor because it does not sense moisture below about 3 kilometers. Second, it doesn’t depict water vapor well at any level, due to cirrus obscuration. Water vapor images are simply maps of infrared brightness temperatures; the data are not in moisture units. The mid- and high-level moisture seen in this image is not an ingredient of heavy rain but often traces the subtropical jet stream. Available over land and water, geostationary water vapor images are available about every 30 minutes at higher spatial resolution than microwave imagery.

Return to Top


  2.6  Total Precipitable Water Question

GOES IR win image of Gulf of Alaska

Here is a GOES-West IR window image of the Gulf of Alaska dominated by a frontal cloud band.

GOES water vapor image of Gulf of Alaska

Comparing the infrared window image to the GOES water vapor image, we notice two things: a moisture plume that we didn't see from the infrared window image, mostly in green, and embedded cirrus in yellow and orange. We've marked the width of the plume aloft in violet. Now, look at the four arrows added to this image. Which blue arrow best estimates the edges of the near-surface moisture?

A.

B.

C.

D.

E. Can't tell from the information given

Arrow E is the correct response. We can't accurately determine
the extent of the plume based on these images alone.

Return to Top


2.7  Total Precipitable Water Feedback

GOES water vapor image of Gulf of Alaska

Arrow D shows the extent of the low-level water vapor, but is was almost impossible to determine the low-level water plume from the two GOES images provided. The water vapor channel is sensitive to high-level water vapor, and the infrared window image depicts the tops of clouds, not water vapor in the atmosphere. Neither shows low-level water vapor.

SSM/I TPW image of Gulf of Alaska

We need TPW imagery, supplied here by the Special Sensor Microwave Imager (SSM/I), to see that the low-level moisture plume is considerably narrower than the high-level moisture plume shown in the water vapor image.

Return to Top


2.8  Cloud Liquid Water Definition

clw conceptual animation

clw conceptual animation
Click to view animation

Cloud Liquid Water (or CLW) is the analog of TPW. It represents the depth of liquid that would be accumulated if all the cloud droplets in a vertical column were compressed into an equivalent depth of liquid water.

Return to Top


2.9  Cloud Liquid Water Example

AMSU Global CLW composite

CLW, like TPW, is expressed in millimeters and is only valid over water. This example shows low to moderate amounts of cloud water in green. In regions of low level convergence, rising motion leads to condensation and larger amounts of cloud water, which appear in red. This AMSU product is valid only up to about .70 mm. Above this, the error begins to exceed accepted values. Research scientists use CLW to assess the properties of stratiform clouds, and forecasters can use CLW to help assess the potential for aircraft icing.

Return to Top


2.10  Cloud Liquid Water Question 1

AMSU global TPW composite

AMSU global CLW composite

TPW and CLW are derived using different measurements and algorithms, but we can make some comparisons. Here are two images of the same scene, both derived from AMSU-B data. One shows TPW and the other shows CLW. The color bars help us quantify the two different parameters being shown.

Question

What conclusions can you draw about the two products from the color bars?
(Choose the best answer.)

a) TPW is about 100 times greater than CLW.

b) CLW is about ten times greater than TPW.

c) The magnitudes of the two fields are about the same.

d) Can't tell from the information given.

The correct answer is a). See the following page for an explanation.

Return to Top


2.11  Cloud Liquid Water Feedback 1

AMSU global TPW composite

AMSU global CLW composite

TPW values are about 100 times greater than CLW. In other words, a typical vertically integrated column of cloudy air contains roughly 100 times more water in vaporous form than in liquid form.

Return to Top


2.12  Cloud Liquid Water Question 2

AMSU global TPW composite

AMSU global CLW composite

Now compare the appearance of the images. The CLW image is less continuous, implying that CLW is more variable, with many regions having zero CLW. On the other hand, the TPW image has more uniform values everywhere, with few values near zero.

Question

Why the difference? (Choose the best answer.)

A. CLW is spotty because the parameter represents upper atmospheric phenomena near the tropopause.

B. CLW is spotty because microwave frequencies do not sense cloud water.

C. CLW is spotty because the nature of cloud cover is by nature spotty, unlike vapor,
which is more continuous in the atmosphere.

D. CLW is spotty because it represents water vapor, which varies over significant
stretches of the ocean.

The correct answer is C. See the following page for an explanation.

Return to Top


2.13  Cloud Liquid Water Feedback 2

AMSU global TPW composite

AMSU global CLW composite

The correct answer is c). Unlike water vapor, clouds are naturally discontinuous in the atmosphere, and there are regions with no clouds, hence, CLW may have values of zero.

Return to Top


2.14  Summary

Return to Top

3.0  Precipitation Estimation Using Microwave Data

3.1  Introduction

AMSR-E rain rate image over GOES IR win image over SE US.

In this section, we examine the scientific basis for satellite estimation of precipitation using microwave remote sensing. After showing imagery and derived-product examples, we will look at other precipitation monitoring techniques. Then we will outline the problem of detecting precipitation against different surface backgrounds, land and sea. In addition, we will cover some of the cloud properties that allow us to estimate precipitation using infrared, passive microwave, and active radar. These concepts will be illustrated using several case examples.

Return to Top


3.2  Microwave and Infrared Imagery Comparison

AMSR-E 89 GHz horizontal  polarization image of New England and the North Atlantic

Precipitation retrievals are more complex than TPW and CLW because they involve satellite measurements of clouds containing both water droplets and ice particles. Here is a raw, 89-GHz AMSR-E image off the northeast coast of the United States. The orange and especially the blue areas offshore mark precipitation associated with a convective frontal system. Data like these are the building blocks of precipitation products.

GOES IR win image of New England and the North Atlantic

The corresponding GOES infrared image shows a more familiar depiction of convective cells along the frontal system.

Return to Top


3.3  Microwave Rain Rate Example

AMSU rain rate global composite

Passive microwave data can be used to infer rain rate and snowfall as shown in this global composite. Unlike CLW and TPW, we can derive passive microwave rain rates and snowfall over both land and sea. Rain rate is expressed in millimeters per hour or inches per hour.

Return to Top


3.3a  Other Precipitation Measuring Techniques

National Radar Coverage

Ground-based radar coverage is good over the United States. Even so, there are coverage gaps shown in grey, and ground-based radar observations are compromised over mountainous terrain. Aside from coastal regions and major islands, oceans have no ground-based radar coverage.

global rain gauge network

The U.S. rain gauge network is dense in spots, though like the radar grid, it has prominent holes. On a global scale, many land areas are relatively unobserved with rain gauges. As with ground-based radar, the oceans have no rain gauge coverage at all.

Return to Top


3.4  Spectral Selection for Precipitation Estimation

electromagnetic spectrim from  visible to microwave wavelengths

infrared  portion of the electromagnetic spectrmicrowave portion of the electromagnetic spectrum

What are the different satellite instruments used to make precipitation estimates? Infrared sensors are familiar to most forecasters, with excellent coverage from geostationary satellites and fine spatial resolution. However, when clouds are present, infrared sensors observe only the temperature of the cloud tops. In contrast, passive microwave sensors on polar satellites observe emissions from water and ice within clouds to produce more reliable quantitative precipitation estimates. Finally, active space-based microwave (or radar) sensors have a role in the overall precipitation monitoring mission, producing the highest accuracy in both the vertical and horizontal dimensions.

Question: Choose True or False

Passive microwave sensing produces more reliable estimates of precipitation than with
infrared methods because of the ability to probe the inside of clouds.

True

False

The correct answer is True.

Question: Choose True or False

Compared with active microwave sensing, passive microwave sensors produce the highest
accuracy precipitation estimates and are capable at providing vertical profiles of precipitation.

True

False

The correct answer is False. Active microwave sensors are capable of producing
more accurate precipitation estimates than passive microwave sensors.

Question: Choose True or False

GOES Infrared precipitation estimates are lower in horizontal and temporal resolution when
compared to microwave estimates.

True

False

The correct answer is False. While microwave sensors produce more accurate precipitation estimates,
GOES-based infrared sensors can produce estimates at higher spatial and temporal resolution.

Question: Choose True or False
Passive microwave sensing of precipitation is less accurate than infrared based methods
since the observed microwave radiation originates mainly from the top of the cloud.

 

True

False

The correct answer is False. Microwave sensors can sense precipitation-sized
hydrometeors in all but the thickest clouds.

Return to Top


3.5  Microwave Frequencies for Precipitation Estimation

microwave portion of the electromagnetic spectrum

Let's look more closely at the range of passive microwave frequencies used for sensing precipitation. The lower frequencies, often referred to as "emission channels," measure precipitation mainly from energy emitted by raindrops. The higher frequencies, or "scattering channels," gather energy scattered by ice particles above the freezing level. Future NPOESS microwave instruments will make use of both the scattering and emission channels for precipitation monitoring and other applications.

Return to Top


3.6  Rain Cloud Structure Model

conceptual model of a convective cloud showing ice hydrometeors above the freezing level

Let's examine a hypothetical convective rain cloud to see how the various satellite observations can be used to quantify precipitation. Above the freezing level, the cloud is dominated by a mixture of small, medium, and large ice particles (or solid hydrometeors). This is the regime where the higher-frequency microwave channels will be most effective in observing precipitation.

conceptual model of a convective cloud showing water hydrometeors below the freezing level

As the larger ice hydrometeors fall through the freezing level, the resulting melting produces falling raindrops (or liquid hydrometeors), and now the lower-frequency microwave channels become the preferred observing tool. Below cloud base the rain drops form a rain shaft.

Return to Top


3.7  Infrared Remote Sensing of Clouds and Precipitation

conceptual model of a convective cloud showing IR energy coming from the top of the cloud

A passive infrared remote sensing strategy for inferring precipitation relies on the observation of energy emitted from cloud tops. Notice that no information about either the frozen or liquid hydrometeors is contained in this upwelling energy; the only information is the cloud-top temperature. This information gives us only a rough idea of the precipitation rate.

e depicting cold cloud tops

Here’s a GOES infrared image off the east coast of the United States that we viewed earlier. We can guess that it’s raining based on the cold cloud tops, but it’s difficult to estimate the actual precipitation rates.

Return to Top


3.8  Microwave Remote Sensing of Clouds and Precipitation: Emission Frequencies

animation of a microwave energy emitted at 37 GHz from a convective cloud
Click to view animation

Now let’s look at how low-frequency channels or emission channels (37-GHz and lower) work. In this case, any information coming from the surface is augmented by emission from liquid hydrometeors and cloud water. Over the ocean, energy leaving the cloud is greater than the surface-based energy entering the base of the cloud from below.

AMSR-E 37 GHz vertical polarization image over Eastern Seaboard and Atlantic Ocean

Here is a 37-GHz image of the area we viewed in the infrared. Compared to the cold ocean background in blue, the atmospheric emission signal from clouds and rain shows up distinctly. By comparison, land is warm and appears deep red.

animation of a microwave energy emitted at 37 GHz from a convective cloud

Over land, emissions from land at 37 GHz have about the same magnitude as emissions from precipitation, making it more difficult to detect precipitation over land.

AMSR-E 37 GHz vertical polarization image over the US Midwest 

In this 37-GHz image over the Midwest, raindrops emit microwave energy, but that energy is difficult to distinguish from the surrounding land area emissions. This is the fundamental limitation of microwave precipitation retrieval over land using the emission channels; the magnitudes of the microwave energy from precipitation and surrounding land are similar.

Return to Top


3.9  Microwave Remote Sensing of Clouds and Precipitation: Scattering Frequencies

animation of a microwave energy emitted at 85 GHz from a convective cloud
Click to view animation

High frequency microwave scattering channels, such as 85 GHz, provide different information than lower-frequency microwave or infrared channels. Upwelling energy comes from the surface, cloud water, and raindrops below the freezing level. However, above the freezing level, the energy is attenuated due to scattering by precipitation-sized ice particles. Thus, the net effect of these large ice particles is to depress brightness temperatures seen by the satellite.

GOES IR win image over the US Midwest

Before we look at an 89-GHz image, here is a GOES infrared image showing cloud tops associated with a frontal system over the Midwest. The dark red color indicates colder cloud tops and likely precipitation, but it is difficult to determine exactly where the precipitation is falling beneath these cloud tops and to determine the rain rate.

AMSR-E 89 GHz vertical polarization image over the US Midwest

Let's switch to the Aqua microwave imager AMSR-E, showing the 89-GHz image for the same area as we viewed with GOES. We can see right through the cirrus we saw on the infrared image to precipitation signatures caused by ice scattering. These blues and greens show frozen precipitation aloft, which will later fall and become rain. While infrared sensors detect only cloud-top temperatures, using the high-frequency scattering channels allows us to observe precipitation cell signatures directly.

AMSR-E rain rate image over the US Midwest

Using the higher scattering channels in the 85 - 91 GHz range, precipitation rates can be derived over land, as shown in this image. The rainfall amounts, shown here for our continuing Midwest case, are calibrated in millimeters per hour.

Question

Which of the following statements is correct? (Choose the best answer.)

a) Higher-frequency microwave channels are less useful over land.        

b) Lower-frequency microwave channels are more sensitive to small and medium
         sized cloud water droplets.

c) Clouds and rainfall are more difficult to observe over ocean than land.

d) Precipitation estimation techniques over land rely more on lower-frequency
         channels because they directly sense small-to-medium hydrometeors

The correct answer is b). Lower-frequency microwave channels are sensitive
to small- and medium-sized water droplets, and because of the uniformly cold
ocean background are capable of producing more reliable precipitation
estimates over water. Over land, energy emitted by cloud water and rain drops
is difficult to discriminate from the relatively warm surface and precipitation estimation
techniques that use higher frequency channels can help due to their sensitivity to cloud ice.

Return to Top


3.10  Active Remote Sensing for Precipitation Estimation

 illustration of active microwave sensing from satellite

Finally, active microwave sensors use a pulse of microwave energy emitted by the satellite. The energy interacts with the atmosphere and clouds, and the signal returned is measured by the satellite. This is like the traditional weather radar, except that the radar is located in space, not on the surface of the earth. In an upcoming section, we'll see an example of precipitation measurement by an active, space-borne microwave sensor.

Return to Top


3.11  Comparing Ground and Satellite-Based Radar with Passive Microwave Data

AMSR-E rain rate image over the US Midwest with NEXRAD regional composite reflectivity comparing cells

Let's compare ground-based and satellite views of the Midwest precipitation. On the left is an active microwave (or radar) image at the same time as a zoom of satellite precipitation rates on the right. Notice that the AMSR-E identifies the same cells seen in the weather radar, but the passive microwave image has a much smoother appearance due to the satellite's lower spatial resolution.

Return to Top


3.12  Summary

Return to Top


4.0  Case Example: Hurricane Ivan

4.1  Hurricane Ivan in the Atlantic

GOES Meteosat composite of Saharan Air Layer imagery

GOES Meteosat composite of Saharan Air Layer imagery

GOES Meteosat composite of Saharan Air Layer imagery

TPW products can tell us about the potential for tropical cyclone intensification, based on the moisture content of the lower atmosphere. We will start with a geostationary Saharan Air Layer (SAL) composite product based on multispectral GOES and Meteosat infrared data. It is not a TPW product, but information about the moisture content of the lower atmosphere can be inferred from the retrievals in the cloud-free areas. The three images track the progress of Hurricane Ivan across the Atlantic. Yellow and orange regions, and especially red and pink regions, indicate a dry air layer, usually of Saharan origin. Suspended dust can contribute to an even stronger SAL signal in dry areas. Blue regions indicate a more moist lower atmosphere. The infrared window channel is overlaid on cloudy regions in black and white gray shades. In these regions, our view of the moisture conditions is blocked. The red, orange, and yellow regions indicate a strong SAL, implying dry conditions where hurricanes tend to entrain dry air and weaken. The blue regions indicate a weak SAL and a more moist lower atmosphere. Here the intensification potential for tropical cyclones is enhanced.

SSM/I TPW composite over the Atlantic

SSM/I TPW composite over the Atlantic

SSM/I TPW composite over the Atlantic

Here is the Atlantic Basin using SSM/I TPW for approximately the same time period. Except where heavy precipitation is present, the product is valid, even in areas of clouds. Note that the color table is reversed compared to the geostationary products. Blue now indicates the intrusion of Saharan air into the Atlantic, and oranges show greater water vapor.

Return to Top


4.2  Comparing Dropsonde and SSM/I Data from Hurricane Ivan

SSM/I TPW composite over the Atlantic

In this SSM/I TPW composite, dry air (shown by blues and greens) has traveled all the way from Africa to affect Hurricane Ivan north of South America. The track marks the location of a reconnaissance flight with dots indicating where dropsondes were deployed to measure the environmental moisture.

1
GPS and Jordan soundings over Caribbean
2
GPS and Jordan soundings over Caribbean

In the dry air north of the storm, marked by blue and green shading in the TPW image, notice the relatively dry layer between 500 and 900 hPa shown by the dropsonde soundings for two points north of Ivan. It fact, it is much drier than the tropical average represented by the "Jordan" moisture profile. Mr. Jordon compiled monthly mean sounding profiles for the West Indies for the ten-year period 1946 to 1955. The Jordon profiles serve as a good approximation to mean summertime conditions in tropical regions with similar climates, and have become a standard for comparison in tropical meteorology studies.

3
GPS and Jordan soundings over Caribbean

Now let's look at a dropsonde location within a more moist air mass. The SSM/I TPW values (shown in orange) indicate that total column moisture is significantly higher than for the previous two locations. The dropsonde profile confirms the presence of a deep layer of moisture extending from the surface to 500 hPa. Compared to the previous two profiles, the bulk of the moisture increase occurred below 500 hPa. Notice that the dropsonde relative humidities are generally higher than the Jordan climatological profile. This is typical for moist conditions within a tropical storm environment.

Return to Top


4.3  Dry Air Interaction with Hurricane Ivan

GOES IR window image of Hurricane Ivan

This series of images demonstrates how dry air entrainment can weaken storms and hurricanes well away from the Saharan source region. Here is an image of Hurricane Ivan before it entered a region of drier air.

GOES IR window image of Hurricane Ivan

After encountering a dry air mass, Hurricane Ivan weakens noticeably.

GOES IR window image of Hurricane Ivan

Two days later, Hurricane Ivan enters a more moist environment, strengthens again, and develops a pronounced eye.

Return to Top


4.4  Summary

Return to Top


5.0  Case Example: Southern California Flooding,
          January 2005

5.1   Introduction

 Rainfall totals over California after the Jan 11 2005 storm

A series of storms dropped excessive precipitation over southern California during the period 6 - 11 January 2005, totaling seven inches in downtown Los Angeles and an amazing 32 inches at Opid's Camp in the foothills of the San Gabriel Mountains. In addition to more conventional datasets, TPW and microwave precipitation rate imagery are used to get recent estimates of precipitation before a series of storms comes into radar range.

Return to Top


5.2  500-hPa Heights over the West Coast and Pacific Ocean

500 hPa geopotential heights over the Pacific near California
Click to view animation

A strong polar jet stream diving south from the west-southwest during this period. Embedded within this moist subtropical flow were a series of disturbances that entered southern California every 24 to 48 hours. A large blocking high shunted storms away from the Gulf of Alaska, a region that has the nickname "the graveyard of storms." Instead of moving into the Gulf of Alaska, storms were deflected around the ridge and roared through southern California. Numerical models sometimes do not forecast precipitation events originating in the subtropics very well. Satellite data, particularly satellite microwave data, is helpful in making more effective and timely forecasts of heavy rain events.

Return to Top


5.3  GOES IR Window over the West Coast and Pacific Ocean

GOES IR window loop over the Pacific near California
Click to view animation

The infrared loop shows us one storm system after another affecting California during this period, but since this is a window channel, clouds are seen rather than vapor. We need to investigate the water vapor environment to assess the potential for heavy precipitation.

Return to Top


5.4  GOES Water Vapor over the West Coast and Pacific Ocean

GOES water vapor loop over the Pacific near California
Click to view animation

The GOES water vapor loop shows vital information about the upper-level flow including interactions between the polar and subtropical jet streams. Areas of subsidence mark the position of the jet stream, seen here as darker grey and orange shading. However, the water vapor channel can't sense the low-level water vapor responsible for feeding the torrential rains over coastal California. We might guess that low-level moisture is arriving from the subtropics, but we can't observe it directly from this imagery.

Return to Top


5.5  AMSU TPW over the West Coast and Pacific Ocean

AMSU TPW loop over the Pacific near the US West Coast
Click to view animation

This animation shows AMSU TPW products over the same period. First one major system, then a second with abundant tropical moisture in red, and then a third system move through the area. Despite the jumpiness of the animation, in many ways it's easier to understand than the GOES water vapor movie we saw earlier. It shows the polar front distinctly as the gradient between blue (or relatively dry air) to the north, and greens and yellows (representing moist air) to the south. This distinct delineation between frontal air masses is an important tool for understanding the synoptic situation. The product is virtually "all weather" over oceans, except where rain (in gray) contaminates the signal. Let's focus on an air mass in the second system, noting tropical moist air shown in red. This plume is the so-called "Pineapple Express" from south of Hawaii and is poised to significantly impact southern California's weather.

Return to Top


5.6  SSM/I TPW Image over the West Coast and Pacific Ocean

SSM/I TPW image of the Pacific of the US West Coast

We can also look at the moisture plume with SSM/I TPW composites to forecast the potential for heavy precipitation. However, high TPW does not tell forecasters where precipitation is occurring. For example, how hard is it raining in the plume of tropical moisture targeting southern California? Let's examine the precipitation rates associated with this plume

Return to Top


5.7  SSM/I Rain Rates over the West Coast and Pacific Ocean - I

SSM/I rain rate composite of the Pacific off the US West Coast

At this point in the storm's evolution, rain rates are still relatively light, mostly 2 millimeters per hour or less, shown as blue pixels offshore. However, it is important to continue monitoring precipitation rates as the storm may intensify over time.

SSM/I rain rate composite of the Pacific off the US West Coast

Twenty-four hours later, a part of the precipitation we have been watching has already moved onshore, but an ominous area of precipitation appears offshore and to the southwest. The rainfall rates in red are near 10 millimeters (0.4 inches) per hour. These are significant rates since they represent an areal average observed over a 25-kilometer field of view. Locally, it might be raining much harder at points within the red pixel area.

During the morning of 9 January, forecasters at the NESDIS Satellite Analysis Branch were tracking this developing swath of heavier rain rates. Their analysis of POES microwave and GOES imagery indicated a large and expanding area of convection, with some localized rain rates between 10 to 18 millimeters (0.4 to 0.7 inches) per hour. Extrapolation of the area's movement also indicated that heavier precipitation could impact the Los Angeles area by afternoon. This information was communicated to National Weather Service Weather Forecast Offices (WFOs), River Forecast Centers (RFCs), and the NCEP Hydrometeorological Prediction Center (HPC) to alert forecasters to the increasing potential for excessive rainfall and catastrophic flooding.

Return to Top


5.8  SSM/I Rain Rates over the West Coast and Pacific Ocean - II

SSM/I rain rate composite of the Pacific off the US West Coast

Two hours later, the area of high rainfall rates can be seen moving toward southern California. At about this same time, National Weather Service forecasters were updating quantitative precipitation forecasts using information from both the latest numerical model output and accumulated satellite guidance. By examining POES microwave-based products similar to what is shown here, NESDIS forecasters observed rates of 30 millimeters (1.2 inches) per hour embedded within the larger area of offshore convection steadily approaching coastal areas. All available guidance suggested a need to increase quantitative precipitation forecasts during the afternoon and evening hours of 9 January across inland areas, and especially for points along the mountain ranges in and around the Los Angeles Basin.

Rainfall totals from 10 Jan 2005 storm for California and Nevada

Later verification, showing extreme accumulations, validated the decision to increase forecasted rainfall amounts. Note the rainfall amount in excess of 6 inches in several locations. Early indications of significant rainfall in the microwave precipitation products gave forecasters valuable information before the storm made landfall.

Return to Top


5.9  Radar Rain Rates over Southern California

NEXRAD composite radar loop over California
Click to view animation

Here is a NEXRAD composite radar loop from about the same time period, late morning on 9 January. It shows precipitation over and off California, but doesn't show the approaching area of elevated rainfall rates in the Pineapple Express, which is out of radar range.

SSM/I rain rate composite of the Pacific off the US West Coast

So we know what's imminent based on the radar, but unless we look at the microwave rates we don't know what's coming beyond radar range. By the time heavy rain is observed offshore by radar it is only a matter of a few hours or less until vulnerable regions are inundated. In other words, radar is a great nowcasting tool, but to provide longer lead times we need to consult the satellite rainfall rates as well.

Return to Top


5.10  GOES Hydroestimator Rain Rates

GOES Hydroestimator loop of the Pacific and  California
Click to view animation

At about the same time during the morning of 9 January, the Hydroestimator, a NOAA-NESDIS precipitation tool, shows the same offshore precipitation we saw in the earlier SSM/I rain rate sequence. The Hydroestimator estimates precipitation rate based on GOES infrared brightness temperatures. It tunes those estimates based on the amount of moisture depicted in numerical model fields, stability parameters, radar, and other factors. The Hydroestimator has the advantage of hourly data refresh rates. The disadvantage is that GOES precipitation rates are dependent on infrared-based cloud-top temperatures that are prone to errors, since the technique does not utilize direct observations of the precipitation. In contrast, microwave precipitation rates are more accurate because they are based on direct observation of emissions from precipitation hydrometeors.

Return to Top


5.11  Precipitation Validation

NCEP stage 4 Rainfall totalf from the 10 Jan 2005 storm for the US

Look at the accumulations by 10 January as the Pineapple Express plume moved ashore. Red and purple shading highlights the deluge in southern California, especially over the mountains. While forecast models can predict events like this in broad terms, and ground-based radar can observe imminent downpours, passive microwave products allow forecasters to build confidence in flood forecasts hours ahead of time.

Return to Top


5.12  Summary

Return to Top


6.0  Case Example: Heavy Precipitation in Alaska

6.1  Introduction

Snowfall totals around Anchorage Alaska for the 16-17 March 2002 storm

NEXRAD Base Reflectivity image over Anchorage Alaska and surrounding area
Click to view animation

A potent winter snowstorm dumped a record 28.7 inches of snowfall on Anchorage Alaska from 16 to 17 March 2002. This major snowstorm was unusual for Anchorage in terms of duration and intensity, with moderate-to-heavy snow falling for 18 consecutive hours, from 10 PM local time on the 16th until 4 PM on the 17th.

While there were complex mesoscale processes at work generating the copious snowfall amounts over the Anchorage and Cook Inlet areas, the extended period of moderate-to-heavy precipitation could not have occurred without a rich supply of moisture.

This case will highlight the role that POES-derived products played in locating the source of the moisture, showing its extent, and in forecasting the duration of significant precipitation across south-central Alaska.

Return to Top


6.2  Large-Scale Synoptic Weather Pattern

GFS 500 hPa heigts and temperatures over North America and the Pacific

During the early morning hours of 15 March 2002, analyses showed a 500-hPa trough positioned near 170 deg W, extending from just west of the Hawaiian Islands northward to the Aleutians. Downstream of that trough, the axis of a high-amplitude ridge stretched north-south along 145 deg W. Like the trough, this ridge also extended from the subtropics to near the Arctic Circle. This had important implications for moisture transport, as we will see in a moment.

Return to Top


6.3  GOES Water Vapor and AMSU Precipitable Water

GFS 500 hPa heigts over North America and the Pacific  over GOES water vapor image
Click to view animation

The short loop shows 3-hour GOES water vapor imagery with 500-hPa heights on 15 March. From the animation it is readily apparent that strong southerly flow and northward transport of moisture was taking place along the western side of this amplified ridge. Notice how the moist flow extended from the tropics northward into Alaska. Since conventional water vapor imagery tells us more about moisture at mid- and high levels, we need microwave products from POES to determine the location and extent of deep layer moisture.

AMSU TPW composite over the Northern Pacific
Click to view animation

The AMSU TPW loop for the same period revealed a formidable moisture plume advecting northward from its origins in the subtropics. From this imagery we also notice that the moisture plume that ultimately fed the developing storm in Gulf of Alaska was initially composed of two separate moisture streams: one originating in the western Pacific and the other near the Hawaiian Islands.

Return to Top


6.4  Numerical Model Depiction and Snowfall Forecast

GFS 700 hPa  specific humidity and 500 hPa heights over the North  Pacific
Click to view animation

GFS 48 hr snow totals over Alaska

During this period, numerical model forecast fields of specific humidity at 700 hPa showed a similar pattern in the moisture field. Two separate moisture plumes combined into one that rapidly expands northward, as the AMSU TPW loop showed. Comparing the two lends confidence to the model’s handling of the moisture field as the storm system takes shape.

In the GFS 84-hour forecast of precipitation, valid at 00 UTC on the 15th, the model already hinted at a potentially significant snowfall event for portions of southwest and south-central Alaska. Large portions of the Kenai Peninsula, Cook Inlet, and Anchorage metro area are within a forecasted area of 8” or greater snowfall totals for the 48-hour period from 12 UTC on the 16th until 12 UTC on the 18th.

Return to Top


6.5  AMSU TPW and Anchorage Radar: Snowfall Begins

AMSU TPW composite over the Northern Pacific
Click to view animaton

The AMSU TPW loop showed the storm evolving over the northern Gulf of Alaska, from 00 UTC on the 16th to the onset of heavy snowfall in the Anchorage area at 07 UTC on the 17th. During this period the moisture plume moved eastward into the Gulf of Alaska, where it fed moisture into south-central Alaska.

Doppler radar loop over Anchorage AK
Click to view animation

Compare the last frame in this loop to ground-based NEXRAD radar between 00 UTC on the 16th and 07 UTC on the 17th. The arrival of the moisture plume coincided very closely with the onset of significant precipitation across the Cook Inlet area and over Anchorage itself.

AMSU composite TPW image ove the North Pacific

Now let’s extend the AMSU TPW loop and focus on the 18-hour period of the heaviest snowfall for Anchorage. Moisture continued to feed into south-central Alaska, supporting further precipitation across the region. Toward the end of the period however, TPW values steadily decreased as the low pressure center over the northern Gulf of Alaska weakened and the axis of strong southerly flow responsible for the moisture advection moved east of the region.

Return to Top


6.6  Summary

Return to Top


7.0  Case Example: Hurricane Katrina Rain Rates

7.1  Introduction

Hurricane Katrina formed as a tropical storm east of Florida, became a hurricane as it crossed the peninsula, and then strengthened into a furious Category 5 storm over the Gulf of Mexico. As a Category 3 storm, Katrina headed inland with a track similar to Hurricane Camille, which devastated the region 36 years earlier.

Visible and infrared composite of Hurricane atrina in the Carribean and over New Orleans
Click to view animation

This loop uses GOES and POES data, infrared at night and visible during the daytime, over a background of natural color during the daytime and city lights at night. But this loop cannot answer one of the important questions that we've been asking in this module: How hard is it raining within the storm?

Return to Top


7.2  Passive Microwave Rain Rates of Hurricane Katrina

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Florida

AMSR-E  rain rate over GOES IR Window  image of Hurricane Katrina off Cuba

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Florida

Passive microwave precipitation products can answer that question in surprising detail. This series of TRMM TMI and AQUA AMSR-E images shows the genesis of precipitation in the storm. After appearing as a disorganized area of convection early on the 23rd, the storm becomes better organized off the Florida coast by the 24th.

Return to Top


7.3  Doppler Radar Rain Rates of Hurricane Katrina over Florida

Doppler radar image of Hurricane Katrina over South Florida

Groud-based radar is an excellent tool to observe the intense precipitation in tropical cyclones. However, it's only useful when storms approach or move over land, for example, in this view of Katrina over Florida. That is why passive microwave precipitation estimates from satellite can be a valuable source of information.

Return to Top


7.4  SSMIS Microwave Rain Rates of Hurricane Katrina over Florida

SSMIS rain rate over GOES IR Window  image of Hurricane Katrina over  Florida

This is the Katrina precipitation rate from the new, advanced Special Sensor Microwave Imager Sounder (or SSMIS) flying aboard the F-16 satellite. It shows rain falling in excess of one inch per hour over South Florida.

Return to Top


7.5  TMI Rain Rates of Hurricane Katrina over the Caribbean

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Florida

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Florida

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Cuba

Using images from the TRMM TMI sensor in a time sequence, we see broadening coverage by green, yellow, and red shades, indicating increasing precipitation rates and storm intensification as the hurricane crosses southern Florida into the Gulf of Mexico.

TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Cuba
TRMM TMI rain rate over GOES IR Window  image of Hurricane Katrina off Cuba
AMSR-E rain rate over GOES IR Window image of Hurricane Katrina in the Carribean

On 27 and 28 August, Katrina strengthened, with sustained winds reaching 105 knots as shown in the AMSR-E image. Amazingly, Hurricane Katrina continued to strengthen into one of the strongest storms ever observed in the Gulf of Mexico.

TMI rain rate over GOES IR Window image of Hurricane Katrina in the Carribean
AMSU rain rate over GOES IR Window  image of Hurricane Katrina making landfall  Near New Orleans

The TRMM TMI shows Hurricane Katrina as a Category 5 storm with winds of 145 knots. Throughout this intense phase, precipitation rates occasionally exceeded 25 millimeters per hour, or over one inch per hour. This AMSU image shows Katrina as it moves onshore.

Return to Top


7.6  Doppler Radar Imagery of Hurricane Katrina over New Orleans

Doppler radar image of Hurricane Katrina near New Orleans

Near land, ground-based radar became available, but forecasters would prefer to get recent estimates of precipitation even before a storm comes into radar range. Plus, sometimes radar precipitation measurements and rain gauges can fail in intense hurricanes, as occurred for some stations when Katrina moved onshore. The need for coverage beyond radar range has led researchers to develop innovative, satellite-based precipitation estimation techniques that will be covered in the following pages.

Return to

top


7.7  Combining Polar and Geostationary Data for Optimal Rain Rate Products

illustration of SSM/I microwave sensor swath width
Click to view animation

Combining polar-orbiting microwave and geostationary data provides opportunities to maximize the advantages of each system. Each polar-orbiting satellite sensor covers a location on the earth every 12 hours. Multiple sensors mean that passive microwave rain rate products for any given location become available every 3 to 4 hours on average.

Illustration of GOES East and West in obrit

Data from geostationary satellites arrives every half hour or even more often, but lacking microwave sensors, geostationary data cannot produce reliable rain rates. Thus, researchers have developed synergistic precipitation products which combine the accuracy of the microwave rain rates with the temporal advantages of geostationary data. These high resolution products have been developed with an eye toward data assimilation, model validation, and climate studies, but they are growing in popularity with operational weather forecasters.

Return to Top


7.8  Experimental Rain Rate Product Examples

NASA blended precipitation product over Hurricane Katrina
Click to view animation


Climate Prediction Center  blended precipitation product over Hurricane Katrina
Click to view animation


NRL blended precipitation product over Hurricane Katrina
Click to view animation

NRL blended precipitation product over Hurricane Katrina
Click to view animation

Here are some experimental rain rate products showing Katrina as the storm moves across the Gulf of Mexico and then makes landfall along the U.S. Gulf Coast. The TRMM Multi-satellite Precipitation Analysis (TMPA) combines microwave estimates of precipitation with geostationary satellite infrared estimates. Microwave rain rates are used to calibrate the GOES estimates. The NOAA CPC Morphing Technique (CMORPH) product is constructed entirely from passive microwave precipitation estimates. At times and locations when polar-satellite microwave data are unavailable, CMORPH propagates the microwave estimates in the time gaps using trends observed in geostationary infrared data. This propagation is referred to as “morphing.” To compute estimates using the Naval Research Laboratory Blended technique (NRL-Blended), passive microwave data from polar-orbiting satellites and TRMM radar data are used to "calibrate" the geostationary infrared data where the microwave and infrared data overlap. This information is retained and used to produce rain rates for continuing calibration of newly received geostationary satellite data. The Self-Calibrating Multivariate Precipitation Retrieval (SCaMPR) algorithm is a rain rate estimation technique that calibrates predictors from GOES data to rainfall rates from microwave instruments. The goal is to produce estimates at the frequency of GOES data but with an accuracy that is closer to that of the intermittently available microwave rainfall rates. SCaMPR is an experimental algorithm that is run in real time but is still under development, and the need for further development is made clear by the underestimation of the heaviest rainfall prior to and during landfall. Click any of the products to see three-hour rain rates for the six-day period from 25 to 30 August 2005.

Return to Top


7.9  TRaP Precipitation Estimation Technique

AMSU rain rate of Hurricane Katrina near New Orleans

To get a short-term precipitation forecast before storms come into radar range, we can utilize the Operational Satellite Derived Tropical Rainfall Potential (TRaP) product. However, since there is a full module explaining TRaP, we will only look at a quick example for Katrina. The AMSU precipitation product just before landfall provides the initialization for the TRaP product. At the time, forecasters were worried about the combination of heavy rains and storm surge in New Orleans. Can this technique offer some help?

Return to Top


7.10  TRaP Example

TRaP product of Hurricane Katrina near New Orleans

This is the TRaP 24-hour forecast based on the previous rain rate product. It advects the observed precipitation forward in time based on the predicted motion of the storm. Notice the swath of torrential rain accumulations (shown as orange shades) in coastal Mississippi. New Orleans is outside the area of heaviest precipitation as indicated by the yellow shading.

AMSR-E 89 GHz rain rate product over the South East US

The validation totals based on this radar and rain gauge blended Stage III precipitation product strongly support the TRaP forecast.

Return to Top


7.11  Summary

Return to Top


8.0  Precipitation Monitoring Missions

8.1  The Tropical Rainfall Measuring Mission

Illustration of the TRMM satellite and the swath width
Click to view animation

The Tropical Rainfall Measuring Mission (TRMM), launched in 1997, was conceived as a satellite mission to study tropical rainfall for climate studies and is the precursor to the Global Precipitation Measurement (GPM) mission covered later in the section. TRMM was quickly recognized as a powerful tool for realtime weather monitoring, particularly for tropical cyclones. As of 2006, it was still going strong. The two most important missions are an onboard precipitation radar (or PR) and the TRMM microwave imager (or TMI).

Return to Top


8.2  TRMM Instruments and Scan Patterns

Illustration of the TRMM PR  and TMI swath and FOV

Let’s take a closer look at the TRMM instruments. The PR is like a weather radar in space, emitting a pulse of microwave energy, shown here in magenta. A return pulse gives accurate information about rainfall rate and structure at a spatial resolution of 5 kilometers and a swath width of 220 kilometers. The TMI microwave radiometer observes in conical pattern either ahead or behind of the satellite across an 850-km-wide swath, shown here in blue. The conical scan pattern is similar to other microwave imagers including the SSM/I and the advanced version of the SSM/I, the SSMIS.

The TRMM PR and TMI work together. The spatially-limited but very accurate precipitation retrievals from the PR are used to calibrate the less accurate but much broader coverage of the TMI.

Return to Top


8.3  TRMM Detects a Squall Line

TRMM PR image of a storm over Oklahoma

This example shows a squall line moving across Oklahoma. The red line through the heart of the squall line marks the position of the vertical cross section shown below.

 


The PR gives a three-dimensional, high-resolution view of precipitation systems from a satellite perspective. The cross section pinpoints the location of the heaviest precipitation, shown in black right at the leading edge of the band.

Return to Top


8.4  TRMM Tropical Cyclone Example

TRMM 85 GHz image of Hurricane Ivan off Florida over a  GOES IR Window

Let's compare precipitation from the TRMM TMI, the microwave imager, and the PR, the precipitation radar. First, we view just the TMI rain rates on top of a GOES infrared window image as it captures Hurricane Ivan on a menacing path toward Florida. Rainfall rates exceed one inch per hour as shown by the red clusters in a rain band near the Florida Peninsula.

TRMM PR  image of Hurricane Ivan off Florida over a  GOES IR Window

Now let’s overlay the more accurate and higher-spatial resolution precipitation rates from the TRMM PR. The PR reveals finer-scale detail. Since it can detect precipitation at a higher spatial resolution, the maximum detectable rates are higher. Notice the extremely high rates in white in the PR.

Return to Top


8.5  Global Precipitation Monitoring

Illustration of current and future Global Precipitation Measurement (GPM) satellite constellation

TRMM was designed to measure precipitation in the tropics, orbiting between about 38 degrees south and 38 degrees north. In the late 1990s and early 2000s, several satellite systems carried microwave missions to monitor and understand global precipitation, and laid the groundwork for the Global Precipitation Measurement (GPM) mission currently underway. Notice the important role of select NPOESS satellites that are planned to have a microwave imaging capability. The central satellite in the GPM constellation is called GPM-Core and will provide calibration-reference for other members of the constellation. This satellite will have a Dual-frequency Precipitation Radar (DPR) and the GPM Microwave Imager (GMI), both improved versions of the instruments on TRMM.

Return to Top


8.6  GPM Dual-Frequency Radar Channels

Graph showing GPM  Radar frequencies and rain rate ranges

The GPM satellite will feature an improved Dual-frequency Precipitation Radar (the DPR) with two sensing instruments: a 14-GHz radar for measurement of relatively heavy rain rates in the tropics and a 35-GHz radar for measurement of precipitation elsewhere, especially at high latitudes. More importantly, the 35-GHz radar will be able to measure light rain and snow. This approach will enable effective measurement of precipitation anywhere on the globe.

Return to Top


8.6a  Flyout Passive Microwave Remote Sensing of Polar Regions

AVHRR IR win image of both poles

Passive microwave measurement of precipitation is difficult in high-latitude regions. In these regions, light rain rates from stratiform clouds over water are difficult to quantify accurately. Over cold land, particularly ice- and snow-covered regions, microwave detection of atmospheric features using many of the microwave instruments currently flying is problematic. Fortunately, higher-frequency channels planned for some of the NPOESS satellites and on all of the dedicated GPM sensors will improve the measurement of precipitation in high-latitude regions.

Return to Top


8.7  GPM GMI Features

Image depicting the Global Precipitation Mesurement (GPM) mission Microwave Imager and labeling its components

The GPM will also carry the conical-scanning GMI, which contains high frequency channels that TRMM TMI does not have. These channels already exist on the modernized SSMIS, also a conical scanner.

 SSMIS 150 GHz  image of Hurricane Katrina

As a preview of GMI capability, here are example SSMIS images centered on Hurricane Katrina. In these images, precipitation is shown in shades of blue. However, these channels are also valuable in the measurement of moisture. The 150-GHz image reveals few traces of atmospheric moisture

SSMIS 183 (+1) GHz  image of Hurricane Katrina

SSMIS 183(+3) GHz  image of Hurricane Katrina

SSMIS 187 (+7) GHz  image of Hurricane Katrina

But notice a significant atmospheric moisture gradient over the central United States in all three of the 183-GHz images. Each channel used to create these images is sensitive to different moisture characteristics. Quantitative moisture profiles are then derived from the raw data for a variety of applications.

Return to Top


8.7a  Flyout: More on SSMIS Channels

graph of the microwave region of the electromagnetic spectrum

The 183-GHz channel is centered in a water vapor absorption band. Sensor developers found that if they expanded the bandwidth on either side of 183 GHz, they could produce channels that were sensitive to different layers of atmospheric moisture. The ±1 GHz channel is the most sensitive to moisture, observing moisture and brightness temperatures within a layer centered near 350 hPa.

Image depicting the 182-GHz plus/minus 3 GHz channel water vapor absorption band

The ±3 GHz channel is somewhat less sensitive to moisture and sees deeper into the atmosphere, observing a layer of mid-level moisture typically centered near 550 hPa.

Image depicting the 182-GHz plus/minus 7 GHz channel water vapor absorption band

The ±7 GHz channel is the least sensitive to moisture and senses a layer of low-level moisture typically centered near 750 hPa.

Return to Top


8.8  GPM Applications: Rain vs. Snow

drawing of the AQUA spacecraft showing the AMSR-EPhoto of the SSMIS instrument

Traditional rain rate algorithms have not performed well in distinguishing falling snow from rain, and there have been of limited use in wintertime situations. However, use of higher-frequency microwave channels, contained on a limited number of instruments, may hold the key to improved snow forecasting.

Drawing of the GMI instrument

Improved resolution on the GPM GMI and future NPOESS microwave instruments will give us the detail that we need to better distinguish rain from snow and the prospect of measuring both from space. This is an important capability that not even current operational Doppler radars can perform. How would the high-frequency channels be able to help us?

Storm total snowfall over the Washington DC area 14-18 Feb 2003

Let’s look at a case study; a near-record snowfall fell across the Washington D.C. area in February 2003.

Return to Top


8.9  GPM Applications: Precipitation Examples

SSMI nad Doppler Radar rain rates over the CONUS
Click to view animation

Prior to the event, we can represent the developing snowstorm using the high-frequency measurements of the the Advanced Microwave Wave Sounding Unit (AMSU-B) on the NOAA satellites. Red means rain, and blue indicates snow. Precipitation is relatively light as the system first moves from west to east, but notice the large area of blue that finally develops over the Mid-Atlantic states. The spatial resolution is relatively coarse using the AMSU-B sensors here, so the precipitation sometimes has a blocky appearance and misses some precipitation compared to the NEXRAD composite to the right. Improved resolution on the GPM GMI and future NPOESS microwave instruments will give us the detail that we need to better distinguish rain from snow, and the prospect of measuring both from space.

Return to Top


8.10  Summary

Return to Top


9.0  Module Summary

Overview

Atmospheric Microwave Products

Precipitation Estimation Using Microwave Data

Case Examples

Precipitation Monitoring Missions

Return to Top


10.0  References

The following resources provide additional information on the topics presented in this module.

Atlas, R., R.N. Hoffman, S.M. Leidner, J. Sienkiewicz, T., W.Yu, S.C. Bloom, E. Brin, J. Ardizzone, J. Terry, D. Bungato, and J.C. Jusem, 2001: The effects of marine winds from scatterometer data on weather analysis and forecasting. Bull. Amer. Meteor. Soc., 82, 9, 1965-1990.

Chang, P.S., P. Gaiser, K. St. Germain, and L. Li, 1997: Multi-frequency polarimetric microwave ocean wind direction retrievals. Proc. International Geoscience and Remote Sensing Symp. 1997 (IGARSS - 97), Singapore, 1009-1011.

Chang, P.S. and Z.A. Jelenak, 2004: NOAA Ocean Surface Wind Vector Retrievals from WindSat Polarimetric Measurements. Presentation, Joint Center for Satellite Data Assimilation (JCSDA) Seminar Series, JCSDA, Camp Springs, MD. [Available online at http://www.jcsda.noaa.gov/seminars/]

Chavanne, C., P. Flament, R. Lumpkin, B. Dousset, A. Bentamy, 2002: Scatterometer observations of wind variations induced by oceanic islands: Implications for wind-driven ocean circulation. Can. J. Remote Sens., 28, 3, 466-474.

Chelton, D.B., M.H. Freilich, and S.K. Esbensen, 2000: Satellite observations of the wind jets off the Pacific Coast of Central America. Part I: Case studies and statistical characteristics. Mon. Wea. Rev., 128, 7, 1993-2018.

Chelton, D.B., M.H. Freilich, and S.K. Esbensen, 2000: Satellite observations of the wind jets off the Pacific Coast of Central America. Part II: Regional relationships and dynamical considerations. Mon. Wea. Rev., 128, 7, 2019-2043.

Cobb, H.D., D.P. Brown, and R. Molleda, 2003: The use of QuikSCAT imagery in the diagnosis and detection of Gulf of Tehuantepec wind events 1999-2002. Preprints, 12th Conf. on Satellite Meteorology and Oceanography, Long Beach, Amer. Meteor. Soc., 410 pp.

Gaiser, P.W. and K.M. St. Germain, 2000, Spaceborne polarimetric microwave radiometry and the Coriolis WindSat system. Proc. IEEE Aerospace Conf., 2, 159-164.

Gaiser, P.W., NRL, cited 2005: WindSat—Remote sensing of ocean surface winds. [Available online at http://www.nrl.navy.mil/content.php?P=04REVIEW87]

Gemmill, W. H. and V. M. Krasnopolsky, 1999: The use of SSM/I data in operational marine analysis, Wea. and Forecasting, 14, 789-800.

Gemmill, W.H., L.D. Burroughs, V.M. Gerald, and V.M. Krasnopolsky, 2000: Ocean Surface Wind Speeds Retrieved from SSM/I on Board DMSP Satellites. NWS Technical Procedures Bulletin, 467, NOAA, U.S. Department of Commerce. [Original available online at http://www.nws.noaa.gov/om/tpb/467.htm; 2004 update available online at http://polar.ncep.noaa.gov/mmab/tpbs/operational.tpbs/ssmitpb/ssmitpb.htm]

Goodberlet, M. A., C. T. Swift, and J. C. Wilkerson, 1989: Remote sensing of ocean surface winds with the Special Sensor Microwave/Imager. J. Geophys. Res., 94, 14,547-14,555.

Goodberlet, M.A., C.T. Swift, and J.C. Wilkerson, 1990: Ocean surface wind speed measurements of the Special Sensor Microwave/Imager (SSM/I). IEEE Trans. Geosci. Remote Sens., 28, 5, 823-828.

Goodberlet, M.A. and C.T. Swift, 1992: Improved retrievals from the DMSP wind speed algorithm under adverse weather conditions. IEEE Trans. Geosci. Remote Sens., 30, 5, 1076-1077.

Jelenak, Z.A., T. Mavor, L. Connor, N. Wang, P.S. Chang, and P.W. Gaiser, 2004: NOAA semi-empirical model functions for WindSat ocean brightness temperatures. Presentation, NRL WindSat Workshop, Solomons, MD, NRL.

Jelenak, Z.A., T. Mavor, L. Connor, N. Wang, P.S. Chang, and P.W. Gaiser, 2004: NOAA WindSat EDR retrievals part II: ocean surface wind vectors. Presentation, NRL WindSat Workshop, Solomons, MD, NRL.

Online SeaWinds on QuikSCAT Documentation, Physical Oceanography Distributed Active Archive Center (PO.DAAC), California Institute of Technology, NASA/Jet Propulsion Laboratory. [Available online at http://podaac.jpl.nasa.gov/quikscat/qscat_doc.html]

Petty, G.W., 1990: On the response of the Special Sensor Microwave/Imager to the marine environment—Implications for atmospheric parameter retrievals. Ph.D. dissertation, University of Washington, 291 pp. [Available from University Microfilm, 305 N. Zeeb Rd., Ann Arbor, MI 48106.]

Petty, G. W., 1993: A comparison of SSM/I algorithms for the estimation of surface wind. Proc. Shared Processing Network DMSP SSM/I Algorithm Symp.,Monterey, CA, NRL.

J. Figa-Saldaña, J.J.W. Wilson, E. Attema, R. Gelsthorpe, M.R. Drinkwater, and A. Stoffelen, 2002: The advanced scatterometer (ASCAT) on the meteorological operational (MetOp) platform: A follow on for European wind scatterometers. Can. J. Remote Sensing, 28, 3, 404-412. [Available online from MetOp-Resources-Publications at http://www.esa.int/esaME/index.html]

Rhome, J.R., 2003: Detecting marine winds from space: An introduction to scatterometry and the current operational scatterometers. Mar. Wea. Log, 47, 2. [Available online at http://vos.noaa.gov/mwl.shtml]

Von Ahn, J., 2003: The impact of QuikSCAT winds on the issuance of marine wind warnings. Mar. Wea. Log, December, 2003, 47, 2. [Available online at http://vos.noaa.gov/mwl.shtml]

Wentz, F.J., 1992: Measurement of oceanic wind vector using satellite microwave radiometers. IEEE Trans. Geosci. Remote Sens., 30, 5, 960-972.

Wentz, F.J., 1997: A well-calibrated ocean algorithm for SSM/I. J. Geophys. Res., 102, C4, 8703-8718.

Return to Top