TOVS Pathfinder C1 MSU Daily pm (Ch 2/3, Ch 4, Ocean Precip)Entry ID: msuc1daypm
Abstract: SECTION 1. GENERAL DESCRIPTION
Spencer and Christy (1990) demonstrated that accurate estimates of
global atmospheric temperatures could be derived from the Microwave
Sounding Units flown on NOAA's TIROS-N series of satellites. The
MSU's have been continuously operating for more than 14 years
collecting measurements of the thermal emission of radiation by
... molecular oxygen at four frequencies near 60 GHz.
The four MSU channels have contribution functions (CF's) determined by
their wavelength and the atmospheric profile of oxygen. The CF's for
the four channels peak succeedingly higher in the atmosphere. MSU
channel 1 has a CF peak near the surface, and is influenced strongly
by temperature and liquid water near the surface. The channel 2 CF is
dominated by a wide peak near 50 kPa. Channel 3 peaks near the
tropopause (25 kPa) and the Channel 4 CF has a somewhat sharper peak
in the lower stratosphere (7 kPa).
Spencer and Christy (1992 a,b) demonstrated that the MSU calibrations
have been very stable, with a precision of monthly satellite
measurements of 0.01 degrees Celsius for the global mean.
A. 2/3 Lower Tropospheric Retrieval.
Spencer has developed a procedure for removing the stratospheric
influence in channel 2 using data from MSU channel 3. The use of
channel 3 allows for a different retrieval at each footprint producing
a robust temperature estimate suitable for daily grid point analysis.
His algorithm (2/3 retrieval) uses a linear combination of channels 2
and 3 to produce a CF with a peak in the lower troposphere near 60
kPa. The retrieval is constrained to reduce the influence of
calibration drifts in channel 3 on several MSU instruments. The 2/3
retrieval algorithm screens the data to remove calibration differences
between satellites along with the effects of precipitation size ice
particles, and corrects for the effect of changing earth
incidence-angle across a scan line (Limb93). The data have been inter
calibrated to correct for differences between local orbit crossing
times (e.g., 2:30 a.m. vs. 7:30 a.m.); however, the ascending and
descending orbits have been processed separately. As a last step, the
algorithm screens out known bad data points found during an exhaustive
analysis of the entire data set.
B. Ch 4 Lower Stratospheric Retrieval.
Spencer and Christy (1993) developed a deep-layer lower stratospheric
temperature retrieval using MSU channel 4. The MSU channel 4 CF lies
nearly entirely in the stratosphere with a peak near 7 kPa. They
validated their 2.5 degree grid point data against rawinsondes,
finding correlation's with single-station data as high as .99. The
stratospheric retrieval corrects the channel 4 data for scan angle
differences across the scan line (Limb93), and removes known bad scan
lines. The data have been inter calibrated to remove calibration
differneces between satellites and to correct for differences between
local orbit crossing times (e.g., 2:30 a.m. and 7:30 a.m.). However,
the ascending and descending orbits have been processed separately.
C. Oceanic Precipitation Data set.
The oceanic precipitation estimates follow the method of Spencer
(1993). Oceanic precipitation is estimated by increased warming in
MSU channel 1 over a threshold. The increased warming is attributable
to emission by liquid water in the lower troposphere. Warming due to
air mass differences are removed using information from the MSU
channel 2/3 retrieval. The precipitation estimate has been calibrated
into mm per day rainfall using eight years of rain gauge data from
tropical island locations. A climatological ice mask has been used to
screen out anomalous precipitation over ice covered ocean. Likewise,
the maximum value allowed has been set at 99.9 mm/day to further
screen out any remaining ice influence.
Also, because this method only works in the Tropics and mid-latitudes,
the retrieval is only calculated for latifudes 60 S to 60 N. Spencer
(1993) discusses the calibration and demonstrates there is good
agreement between the MSU precipitation estimates and previous climate
rainfall data sets. Because the algorithm methodology uses all orbits
in a 24 hour period to calculate the rainfall estimate, no diurnal
information is available in this data set.
D. MSU Calibration.
All the Pathfinder C1 retrievals use calibrated MSU data. The MSU
brightness temperatures for all four channels were linearly calibrated
using the radiometric counts measured viewing the earth scene as
compared to the difference between the radiometric measurement while
viewing a warm calibration load and a similar view of deep space
(assumed to be 2.5 K).
SECTION 2. TOVS PATHFINDER C1 DATA FILE STRUCTURE
This file contains TOVS Pathfinder C1 data averaged from the
p.m. orbits for a single day.
The daily data product file contains data for seven TOVS level 3
geophysical parameters given in separate scientific data sets (SDS's)
in HDF format. The first scientific data set in each file contains
the lower troposphere deep- layer mean temperature (Ch 2/3) estimate
derived using Spencer's retrieval method. The second SDS contains the
lower stratosphere (Ch 4) deep-layer mean temperature estimate
(Spencer and Christy, 1993) and the third SDS, the oceanic
precipitation estimates from MSU (Spencer 1993). The next two SDS's
(four and five) contain the grid point standard deviations for the Ch
2/3 and Ch 4 parameters, while the last two (six and seven) contain
the grid point sample count parameter for the Ch 2/3 and Ch 4
products. The data sets have been processed by the Observing Systems
Branch of the Space Science Laboratory at NASA/MSFC, using the
algorithms described in Section 3 (For complete details of the
processing algorithms, the user is referred to the TOVS Pathfinder C1
User's Guide or the references listed in Section 4.).
The standard deviation and counts provided in the Path C1 data files
are calculated from data for all the footprints assigned to each grid
box. Details of the grid box assignment and weighting are discussed
in the TOVS Pathfinder C1 User's Guide. The counts and standard
deviations are provided for the two deep-layer temperature products,
but not for the oceanic precipitation.
SECTION 3. ALGORITHM DESCRIPTION
A. Lower Troposphere Deep Layer Mean Temperature.
SDS 1 contains the lower tropospheric temperature parameter (Ch 2/3)
which is derived using a linear combination of MSU channels 2 and 3:
2/3 Tb = 1.6*Tb2 - .6*Tb3
where, Tb2 refers to the brightness temperature in MSU channel 2 and Tb3 the
brightness temperature in MSU channel 3.
The 2/3 retrieval is calculated for each footprint of the MSU scan
line, the latitude and longitude of the footprint are calculated and
the appropriate limb correction (Limb93) is applied based on latitude,
longitude, month, and scan angle. Next, the footprint data are
assigned to adjoining grid boxes in a weighted method depending on
footprint latitude. After all the p.m. orbits are processed, a
horizontal averaging is used to fill some of the empty grid boxes.
Once the final 2.5 degree grid is completed, the data are resampled at
a 1 degree grid spacing.
B. Lower Stratosphere Deep Layer Mean Temperature.
The lower stratospheric temperature (Ch 4) is given in SDS 2. The
stratospheric temperature retrieved is simply the black body
temperature linearly calibrated from the radiometric counts in MSU
channel 4 at each footprint.
The stratospheric retrieval is calculated for each footprint of the
MSU scan line, the latitude and longitude of the footprint are
calculated and the appropriate limb correction (Limb93) is applied
based on latitude, longitude, month, and scan angle. Next, the
footprint data are assigned to the appropriate grid box. North
(south) of 66.7 degrees N (S) the footprint data are assigned to
adjoining grid boxes in a weighted method depending on footprint
latitude. Section 3 describes the grid box averaging. After all the
a.m. orbits are processed, a horizontal averaging is used to fill some
of the empty grid boxes. When the final 2.5 grid is completed, the
data are resampled at a 1 degree grid spacing.
C. Oceanic Precipitation.
The methodology used to calculate the oceanic precipitation is
highlighted in Section 1 C. For a more in-depth discussion the reader
is referred to Spencer (1993). Because of the algorithm design, the
a.m. and p.m. orbits are not separated, but are combined into a daily
rainfall estimate, so no analysis of a diurnal effect is possible.
Likewise, because of the algorithm design, the standard deviations and
grid box counts are not available for this product. The precipitation
estimates are calculated on a 2.5 degree grid (for latitudes 60 S to
60 N) and are resampled onto the 1 degree grid. The Path C1 data
files contain the precipitation estimates on the 1 degree grid, with
the same rainfall product provided in both the a.m. and p.m. data
Data Set Citation
Dataset Originator/Creator: Spencer R., Christy J.R.
Dataset Title: TOVS Pathfinder C1 MSU Daily pm (Ch 2/3, Ch 4, Ocean Precip)Online Resource: http://mirador.gsfc.nasa.gov/cgi-bin/mirador/presentNavigation.pl?p...
Start Date: 1987-03-01Stop Date: 1988-11-30
Latitude Resolution: 2.5 Degree
Longitude Resolution: 2.5 Degree
Horizontal Resolution Range: 250 km - < 500 km or approximately 2.5 degrees - < 5.0 degrees
Temporal Resolution: Daily
Temporal Resolution Range: Daily - < Weekly
Distribution Media: On-line
Spencer, R.W., 1993: Global oceanic precipitation from the MSU during 1979-92
and comparisons to other climatologies. J. Climate, 6, 1301-1326.
Spencer, R.W., and J.R. Christy, 1993: Precision lower stratospheric
temperature monitoring with the MSU: Technique, Validation, and
Results 1979-1991. J. Climate, 6, 1194-1204.
... Spencer, R.W., and J.R. Christy, 1992a: Precision and radiosonde validation of
satellite grid point temperature anomalies, Part I: MSU channel 2.
J. Climate, 5, 847-857.
Spencer, R.W., and J.R. Christy, 1992b: Precision and radiosonde validation of
satellite grid point temperature anomalies, Part II: A tropospheric
retrieval and trends 1979-90. J. Climate, 5, 858-866.
Spencer, R.W., and J.R. Christy, 1990: Precise monitoring of global
temperature trends from satellites. Science, 247, 1558-End
Extended Metadata Properties
(Click to view more)
Creation and Review Dates
Last DIF Revision Date: 2014-08-11