University of New Hampshire
  Institute for the Study of
  Earth, Oceans, and Space

in collaboration with:
  Michigan Aerospace Corp.
  University of Hawaii
  Mt. Washington Observatory


. . .
The troposphere is the region of the atmosphere that is closest to the Earth’s surface where clouds and most weather-related phenomena occur. Winds in the troposphere, in fact, have a dominant effect on weather. Measuring tropospheric winds is crucial to the understanding of atmospheric and climate dynamics for weather forecasting. Presently, wind data are collected mainly from a worldwide network of weather-balloon launches; but most of the globe (specifically, over the oceans) is not covered by this network. This sparse input is one of the factors that limit the accuracy of the forecast models. The objective of the GroundWinds program is to develop and demonstrate remote-sensing technologies that can be used to measure global tropospheric winds from orbiting satellites. The resulting data would be used as input to weather modeling calculations, with the goal of greatly improving the accuracy of long-term (greater than 2-day) weather predictions.

GroundWinds is a NOAA-funded initiative to demonstrate the capability of “Direct Detection Doppler LIDAR” to measure wind velocities. LIDAR is an acronym for “Light Detection And Ranging.” GroundWinds uses pulses of light from a laser, which are directed out into the atmosphere and scatter off of the gas molecules, as well any particles or droplets (aerosols), in the air. The light that is scattered back passes through an optical and electronic system that looks for changes in the wavelength of the light (Doppler shift) caused by bulk movements in the atmosphere, namely the wind. “Direct Detection” refers to the way in which the Doppler shift is measured.

The LIDAR system makes use of a Fabry-Pérot interferometer as a high spectral resolution element, capable of detecting Doppler shifts of the backscattered signal that correspond to velocities less than 1 m/s. A schematic that illustrates the overall concept is presented below.


Image used by
permission of
Michigan Aerospace Corp.



An idealized version of the spectrum of the processed return signal is shown. This spectrum is recorded as a function of time to deduce Doppler shift of the light as a function of altitude. It should be noted that this spectrum consists of two distinct components. These are a broad distribution that is caused by “Rayleigh-Brillouin” scattering from atmospheric gases, and a narrow distribution that is the result of “Mie” scattering from atmospheric aerosols. The GroundWinds system utilizes both of these components to produce a measurement of wind velocity. Each of these spectral components is measured using an interferometer whose resolution is optimized for that particular measurement.

GroundWinds LIDAR facilities have been established in Bartlett, New Hampshire and at Mauna Loa, Hawaii. These facilities are being used to test certain technologies that are important to the success of the “DDD” LIDAR technique. In addition to these ground-based “test beds,” a LIDAR system called “BalloonWinds” is being developed, which will be carried above the troposphere under a high-altitude balloon for flights lasting a matter of some hours. These will demonstrate how LIDARs can work from locations closer to the viewpoint of Earth-orbiting satellites.

. . .

The key elements of GroundWinds are: a) the laser, b) the transmitting telescope, c) the receiving telescope, d) the interferometer, and e) an instrument control and data processing system. The laser beam exits the instrument through a beam expander, which reduces the divergence angle of the emerging light to match the telescope field-of-view. The laser light beam is scattered by the molecules and aerosols in the atmosphere resulting in a detectable backscattered light signal. This backscattered light is collected by the primary receiving telescope. A fiber optic cable carries the scattered light from the main receiving telescope to the interferometer subsystem.

Image used by permission of Michigan Aerospace Corp.


After some filtering, the light is introduced into the molecular interferometer. The circular fringe pattern that is created by the molecular interference optics (etalon) is present in the transmitted light, the reflected light from the etalon being the complement of the transmitted fringe pattern. In most interferometers, the reflected light is lost from the system and represents a significant inefficiency. However, in the GroundWinds interferometer the reflected light is re-injected into the molecular etalon in a process called “photon recycling.” Fiber optic arrays are used to collect light that is reflected from the etalon and to reintroduce it into the interferometer. This can be done a number of times to pass more light through the instrument.

After a couple of recycling stages, the primary and recycled light is focused through a “Circle to Line Interferometer Optical” system (CLIO). This innovative device converts the circular fringes from the Fabry-Pérot into a linear pattern that is detected by a charge-coupled device (CCD) camera. The remaining reflected light is injected into the aerosol interferometer where reflected light again goes through recycling. Here again an etalon creates a velocity-sensitive circular fringe pattern that is transformed into a linear pattern by a CLIO coupled to a CCD camera.

The light that was reflected on every injection is passed to a photomultiplier tube (PMT). The PMT is used to measure a photometric intensity profile. This can be used to quantify the integrated energy returned, and has value in numerically correcting for any misalignment in the etalons. The system detectors record at any one time three sets of information; 1) the aerosol fringe pattern optimized to detect the motion of aerosols, 2) the molecular direct fringe pattern, optimized to detect the motion of the molecular component of the atmosphere, and 3) the integrated photometric return.



Image used by
permission of
Michigan Aerospace Corp.


This picture shows how the circular pattern of the Fabry-Pérot fringes are transformed by the CLIO optics into a linear pattern of “spots” that is focused onto a CCD. The outer fringes are produced by the light that passes through the recycling fiber optics. This allows additional “orders” of the fringe pattern to be analyzed for Doppler shifts. As the light from a laser pulse returns through the atmosphere, the spot pattern is shifted through the CCD, forming a group of “streaks.”


Image used by
permission of
Michigan Aerospace Corp.



The distance along the length of the streaks is proportional to the distance from the LIDAR to the region of the atmosphere that scattered the light pulse. The deviation of the “spine” of the streak from a known reference position is a function of the Doppler shift caused by the wind velocity in that region. The upshot is that the wind vector at points along the line of sight (LOS) of the LIDAR can be measured. Combining this with vectors observed in different directions yields a sample of the wind velocity field surrounding the LIDAR. In addition, the width of the spine gives information about the temperature of the air along the LOS.

. . .

Every camera image is a time-resolved measurement of the CLIO output accumulated over many laser pulses. Each image contains a reference region, where the spectral information of the out-going laser is recorded, and a sky region, which contains the spectral information of the backscattered signal. The streak rate of the camera and the speed of light are used to interpret the time-resolved record as a range-resolved measurement, while the look-angle of the laser-telescope determines the range-altitude relationship. In this manner, each row in the camera image is associated with an altitude range and the corresponding row counts contain the spectral information of the laser light backscattered at that altitude. The standard streak rate and number of rows used with the LIDARs in New Hampshire and Hawaii correspond to a maximum range of 25 km and range bins no smaller than 110 meters.

Image used by permission of Michigan Aerospace Corp.

The data undergoes frequency and intensity calibrations. The multiple interference orders are used for frequency calibration purposes (determination of the frequency-pixel column relationship).  Standard flat fielding techniques are used to calibrate the intensities of the measured row counts into meaningful spectra. The calibrated spectrum resulting from the reference region contains important information on the laser spectrum and the instrument response. The altitude-resolved sky spectra are interpreted through the use of the reference spectrum, molecular (Rayleigh-Brillouin) scattering model, aerosol (Mie) scattering model, and background models. Specifically, the sky spectrum from each altitude is fitted to these models.  The fit parameters are Doppler shift, molecular signal strength, aerosol signal strength, and temperature (optional). The algebraic manipulations of these fit parameters produce the primary data products shown below.

The vertical scale of each plot is the altitude in km above sea level. The horizontal scale is time in minutes, but each band of data represents a 30 degree clockwise sweep in azimuth. The blank sections are times when the azimuth is advanced by a large step and the receiving fiber optic is realigned.

Component of the wind velocity along the line of sight
as measured by the "molecular channel" of the LIDAR.



Intensity of the LIDAR return signal measured by the
molecular channel in terms of the digitized CCD camera output.



The aerosol scattering ratio is a measure of the relative strength
of the return signals from aerosol and molecular scattering.
A strong aerosol return is evident here between 8 and 10 km ASL.

Air temperature can be deduced from the width of the
fringes in the molecular channel.


These measurements along with other atmospheric models and theories give rise to secondary
data products, which include:

The horizontal wind speed in units of meters per second
as a function of altitude.


The wind direction as a function of altitude.
Zero or 360 degrees implies a wind from the north.


Optical properties of the aerosols causing the scattering can be deduced
from LIDAR data. The measured aerosol backscatter coefficient (dashed line)
and a molecular backscatter coefficient calculated from scattering theory
(dotted line) are combined to give a backscatter total.

. . .

Institute for the Study of Earth, Oceans and Space
University of New Hampshire

39 College Road, Durham, NH 03824-3525
Tel: 603-862-2867

UNH Project Staff
  Berrien Moore III Principal Investigator (
  James Ryan Co-Principal Investigator (
  Ivan Dors Co-Principal Investigator (
  John McHugh Investigator (
  Philip Dunphy Research Scientist (
  David Bodet
Research Project Engineer
  Colin Frost Research Project Engineer (
  Brian King Research Project Engineer (
  Terry Richardson Project Manager (
  Christian Seymour Chief Operator (