Bruce L. Gary, 2001 September 8

This web page describes the status of an ongoing measurement program using existing JPL radiometers for the study of stratus cloud formation and dissipation.

The observing program is called ROOF, for Radiometric Observations of Fog.  The goal of ROOF is to evaluate the merits of further pursuit of the idea that the formation and dissipation times of West Coast airport fog and stratus clouds can be predicted using a specific complement of microwave remote sensors.  The sensor system consists of two microwave radiometers and various support sensors.

The two radiometer instruments consist of a water vapor radiometer, WVR, and microwave temperature profiler, MTP.  The WVR is the D-01 unit, operating at 20.7 and 31.4 GHz.  The MTP is a slightly modified version of the airborne MTP/ER2 (built and used in NASA's ER-2 aircraft).  The modifications consist of a ground based mounting and a change in frequencies.  The MTP operates at 54.00, 55.47 and 58.80 GHz.  A fuller description will be given in a later section.

The ROOF site is described on a web page of photographs of the microwave radiometers, where they are set-up, and support sensors.  The following picture is from that web page.

Microwave Remote Sensing Instruments

The MTP is a ground-based mounted version of the airborne ER-2 MTP.  It scans through a different sequence of elevation angles and cycles through three frequences (instead of two) at every scan location.  The frequencies are 54.00, 55.47 and 58.80 GHz.  There is a hot target in the nadir direction, and an ambient target (usually about 15 K warmer than outside air ambient) at the northwest horizon location.  Sky readings are taken along the southeast azimuth, from the horizon to zenith.  The observing sequence is: nadir (hot), horizon, horizon, 5.7, 9.0, 14.4, 23.3, 39.0, 90, 90 degrees, ambient target plus noise diode, ambient target (without noise diode on).  A data cycle takes 20 seconds.  Data are recorded in the system's data unit, on the roof, and a serial output of raw data is recorded by a notebook computer inside the residence.  At 24-hour intervals the notebook's data file is transferred to a desktop for analysis.

The WVR is the D-01 unit, operating at 20.7 and 31.4 GHz.  Elevation scans are along the southeast/northwest great cricle azimuth line.  Both channels of data are recorded at each of the following elevation angles (along the southeast azimuth):  30, 41.8, 90, 138.2, 150, 138.2, 90, 41.8, 30, ambient target.  Real-time analyses are performed by the controlling computer (inside the residence), and graphical displays are presented of the following tip curve results: TBzen20.1, TBzen31.4, TBzen31.4(inferred from 30 degree elevation), path delay and zenith integrated liquid water content.

Supporting meteorological data include three ambient air temperatures (one StowAway in the orange tree, a manually logged Radio Shack digital sensor in white box, manually logged Taylor digital sensor in another white box), two pressures (one StowAway in the orange tree, and a manually logged hiker's watch barometer), an IR radiometer that is manually logged, Meteogram data from the Santa Barbara Airport (4.3 miles to the west-southwest, at 70 feet altitude), radiosondes from Vandenberg, VBG (50 miles to the west-northwest), coastal wind profiler (915 MHz, GLA, 5 miles to the west near the coast, high altitude coverage mode and low altitude coverage mode), surface properties near the coast (solar irradiance, pressure, temperature, dew point temperature, wind speed, wind direction).  The Meteograms contain cloud cover, ceiling, barometric pressure, temperature and dew point temperature at 1-hour intervals.  The coastal surface data have better than 1-hour resolution.  The wind profiler data have 1-hour resolution.

Since the instruments are at my residence, I make frequent records in the observing log of two air temperatures, pressure, cloud cover, estimated ceiling, and miscellaneous other conditions (such as when I wipe dew or mist off the radomes).

Calibration of WVR Radiometer

The purpose of the WVR radiometer is to measure zenith integrated content of cloud liquid, i.e., "cloud liquid."  Two effects must be corrected for: 1) water on the radome, caused by either an accumulation of dew or mist, and 2) changes in the altitude distribution of water vapor.  The first effect produces large increases in TB for both channels, which has the effect of producing solutions for cloud liquid that are too high.  The second effect causes changes in cloud liquid zero offsets.  The first effect is overcome by avoiding use of the zenith view whenever there is the possibility of radome water, since radome water often accumulates at the zenith location but not on the sides.  The second effect is overcome by deriving values for a parameter related to an equivalent e-folding altitude for the water vapor height distribution.  This is done during clear times, and since this parameter varies with synoptic time scales, it can be interpolated through the cloudy times.  Details of these two calibrations can be found at WVR Calibration.  The following figure is a version of zenith cloud liquid burden that includes calibration corrections for both the radome dew and water vapor altitude distribution effects.  Comments on this data are deferred to a later section.

Comments on this cloud liquid data are found at Lz Comments, where a discussion is presented of Lz-values sufficient to produce mist and obscure the sun.  A meteorogical-based use of the Lz data will be found in a web page linked to below.

Calibration of MTP Radiometer

The MTP radiometer must undergo an elaborate calibration whenever it is modified.  Since we are using a new set of frequencies, having unkown gain properties, and since the instrument is mounted in a new box with a new radome, with unknown absorption and reflection properties versus viewing angle, these aspects of the MTP instrument must be determined.  The lengthy procedure for calibrating the MTP is presented at another web page: MTP Calibration.

In this section I reluctantly conclude that the MTP cannot be adequately calibrated, and any attempt to produce T(z) histories for the study of stratus formation and dissipation using the ROOF data set is not justified.

Analysis of Support Data

A descrription of this section is not justified, given that the MTP cannot be adequately calibrated.

Interpretation of All Data

First Impressions

The most striking thing about this data set is the low values for cloud liquid burden that were measured by the WVR.  Shadows were lost for burden values of approximately 30 microns, and mist precipitation occurred in association with a burden of about 100 microns.  All my previous WVR measurements of stratus clouds produced much higher values.  For example, based on measurements at JPL, Pasadena's Eaton Canyon Nature Center, El Monte Airport and Ontario Airport, I derived the "rule of thumb" that as stratus clouds formed or dissipated a liquid burden of 100 microns was the criterion for the transition between shadows and no shadows (if it were daytime), and 700 microns was the transition value for mist precipitation onset or cessation.  All of these sites are "inland" - ranging from 24 miles from the coast (JPL and El Monte Airport) to 35 miles inland.  The ROOF site, on the other hand, is only 2.5 miles from the coast.  It is typical for there to be a strong gradient of coastal stratus "presence" over the first inland mile or so of Santa Barbara coast.  The data suggest, to me, that several properties of coastal stratus vary with distance inland.  To account for the low burdens at the ROOF site, it is necessary to hypothesize that the drop size distribution varies with distance inland.  I'm getting ahead of myself here, but this much can be speculated upon by a casual look at the plot of cloud liquid burden for the ROOF observing period.


WVR measurements are self-calibrating due to the use of "tip curves."  MTP measurements are not so easily calibrated.  Either an extensive pre-deployment calibration should be performed, involving the establishment of a tTGT offset (using the ambient pyramid horn) and hot target emissivities using a heated pyramid horn; or, preferably, the MTP should be calibrated at a radiosonde launch site before being deployed to the stratus observation site.  The use of an IR radiometer is hampered when the viewing horizon has non-atmospheric objects within the applicable range of the MTP channels.  Any field calibration of an MTP is hampered when it is likely that horizontal gradients are present.  Such horizontal gradients are not show-stoppers for the production of T(z) profiles, provided the MTP has been calibrated before use at the site with horizontal gradients.

The ROOF observations had just this one fatal flaw:  the MTP could not be calibrated at the ROOF site, nor were calibrations performed after the ROOF observations (since the MTP was needed for another project shortly after ROOF).

In the future, I recommend 1) use of the ambient horn for establishing a tTGT offset, and 2) deploying the MTP to a RAOB site for several days as a first phase of the field observations for establishing channel gains.


Link to previous studies of stratus cloud formation and dissipation.

This site opened:  May 31, 2001 Last Update:  September 8, 2001