Engineering topics
Publications and source records attributed to Kakar, R. K..
The 13.8 GHz Precipitation Radar (PR) aboard the US/Japan Tropical Rainfall Measuring Mission (TRMM) satellite is the first rain profiling radar ever launched into space. A TRMM follow-on mission, called the Global Precipitation Mission (GPM), is currently planned to extend and improve the TRMM acquired rainfall data set.
Explore the source record for details and available documents.
The algorithm developed by Wilheim (1990) dealt with the retrieval of water-vapor profiles from microwave radiometric measurements, even in the presence of clouds. This algorithm was tested using radiometric measurements at frequencies near 90 and 183 GHz; the results show that the retrieved water-vapor profiles were in general agreement with those derived from radiosonde data. In particular, retrieval over both land and ocean surfaces generated clouds at locations that could be verified from satellite photographs. The algorithm did not perform well where there were surface fronts. These clouds possibly contained ice particles as well as precipitating snow, which the algorithm was not expected to handle.
Rapid linear algorithm retrieves meteorological parameters from microwave radiometric data. Versatile 2-step, multiple-linear-regression algorithm examines all possible subsets of available channels and selects optimum subset for retrieving given geophysical or meteorological parameter.
A multiple linear regression technique has been applied to retrieve continuous sequences of atmospheric moisture profiles from a set of measured data. In this method the selection of an optimal subset of sensor channels plays a crucial role, in order to reduce the impact of data noise and redundancy. The data were obtained with a 4-channel remote-sensing microwave instrument carried aboard an aircraft. In contrast, most previously reported moisture profile retrievals from microwave radiometry have been used on simulated, discontinuous data. Although the data were obtained over a land surface with only a limited amount of correlative data, the retrievals were quite successful.
A statistical correlation technique is applied to the retrieval of vertical moisture profiles from downlooking radiometric measurements of atmospheric radiation at microwave wavelengths. Only an optimum subset of available radiometer channels is selected for estimating water vapor at specific pressure levels. To test its validity the algorithm was applied, in a numerical experiment, to 50 independent tropical radiosondes over a sea surface. It was also used to retrieve continuous sequences of atmospheric moisture profiles from a set of data obtained with a 4-channel microwave radiometer carried aboard an aircraft over a land surface.
A statistical correlation technique is applied to the retrieval of vertical moisture profiles under clear-sky conditions from down-looking radiometric measurements of atmospheric radiation at microwave wavelengths. For a given set of channels, the method selects the optimum radiometric channels for estimating water vapor at specific pressure levels between the surface and 300 mb. The water vapor mixing ratio at these pressure levels is then calculated from a linear combination of the selected channel brightness temperatures. To test its validity the algorithm was applied, in a numerical experiment, to fifty independent tropical radiosondes. The rms absolute deviation of the estimated moisture profiles from the actual profiles was comparable to that obtained using an iterative retrieval method reported earlier. The statistical method, however, requires several orders of magnitude less computer time than the iterative method; it is suitable for high speed processing of large amounts of data.
A technique for retrieving vertical moisture profiles from downlooking radiometric measurements of atmospheric radiation near the 183 GHz water vapor line is described. A simulation experiment utilizing this retrieval technique and temperature and moisture profiles from tropical radiosonde stations has been carried out. Assuming the atmospheric temperature profile is known to 2 K rms (state-of-the-art with present remote sensors) and the instrument noise is 1-2 K rms brightness temperature (achievable with current technology), the humidity between the 700 and 1000 mb pressure levels was retrieved to within about 20 percent of true value over ocean and within about 45 percent of true value over land. The precipitable water content was retrieved to within about 10 percent rms of true value over the ocean and within about 45 percent rms of true value over land.
A linear statistical technique using a 'leaps and bounds' procedure (Furnival and Wilson, 1974) is developed for retrieving geophysical parameters from remote measurements. It is used for retrieving sea surface temperatures from the Scaning Multichannel Microwave Radiometer (SMMR) on Seasat. The technique uses an efficient algorithm to select the best fixed-size subset of the 10 SMMR channels for linearly retrieving a given geophysical parameter. The 5-channel subset (6.6V, 6.6H 10H, 18V, 21H), where V and H refer to, respectively, the vertical and horizontal polarizations and the numbers are the channel frequencies in gigahertz, gives the minimum rms error in estimating the sea surface temperature. A comparison with ground truth indicates that the algorithm infers the temperature with an rms accuracy of better than 1.5 K under most environmental conditions. A quality control procedure which is seen as holding promise for further improving the accuracy is proposed.
Numerical experiments using regressions by leaps and bounds have been performed to determine the optimum frequencies for satellite-borne microwave radiometers to estimate atmospheric electrical path length over the sea. The frequency range 5-40 GHz was searched. The effect of surface wind speed, sea surface temperature, and clouds was considered in the optimum frequency selection. The analysis indicates that approximately 0.6-cm rms accuracy is possible for one-way path length measurement using a proper pair of frequencies. The best two-channel subset selected by the leaps and bounds techniques is (16.0, 21.0) GHz.
A model for computing microwave emissivity of a wind-driven foam-covered sea is presented. The effect of roughness and foam is modeled by combining early measurement results and theoretical analysis. Recent Seasat-SMMR measurements are used to fine tune the model and derive an 'effective' fractional foam coverage expression in terms of frequency and wind speed. The model incorporates polarization characterization and view angle dependence of the foam-cover emissivity. For 48.8 deg incidence angle and wind speed less than 15 m/s, the emissivity values calculated from this model differs by less than 8 percent from those calculated from Wilheit's model. At nadir and at 25 m/s wind speed, the emissivity calculated from the two models differ by approximately 15 percent.
Observations of the 183-GHz rotational transition of water vapor in interstellar molecular clouds are reported. The observations were made with a portable double-sideband superheterodyne radiometer used with the 91-cm Cassegrain telescope on board the Kuiper Airborne Observatory. An emission feature was detected in the direction of the Kleinmann-Low nebula in Orion with a peak antenna temperature of 15 K, a local standard of rest velocity of 8 km/sec, and a width of 15 km/sec. A plateau component of the emission profile is attributed to the 1-arcmin region characteristic of plateau emission from other observed molecules, with emission enhanced over that expected for thermal excitation, while the spike component observed is consistent with an optically thick source of the size of the molecular ridge in Orion at a temperature of 50 K and a column density greater than or equal to 3 x 10 to the 17th/sq cm, implying that H2O is one of the more abundant species in the Orion Molecular Cloud. Emission at 183 GHz was not detected in Sgr A, Sgr B2, W3, W43, W49, W51, DR 21, NGC 1333, NGC 7027, GL 2591 or the Rho Oph cloud; it may have been detected in M 17.
Explore the source record for details and available documents.
Three millimeter-wavelength transitions of ClO were searched for in the stratospheric emission spectrum during May-August 1977 using aircraft-based instruments. The measurements indicate that if stratospheric ClO has a vertical distribution of the form predicted by present photochemical models, then the ClO volume mixing ratio at the profile peak did not exceed 10 to the -9th at the time and place of the measurements. The observed spectra suggest the presence of ClO in the stratosphere but do not provide a definitive detection.
Explore the source record for details and available documents.
The reported observations of the J = 0-1 rotational transition (115271.2 MHz) of carbon monoxide in the Venusian atmosphere indicate that the distribution of stratospheric CO is variable. Observations were obtained with a 11-m antenna on several days in February 1977, when Venus was near maximum eastern elongation (k approximately 0.4), and in April 1977 during inferior conjunction (k approximately 0.01). Differences in the line intensities and shapes for the two data sets are described. The results are interpreted to suggest that on the nightside of Venus there is relatively more CO at higher altitudes and relatively less CO at lower altitudes than on the dayside of the planet.
The 115-gigahertz microwave line of carbon monoxide has been detected in the spectrum of Mars. The measurement is sensitive to carbon monoxide between the surface and an altitude of approximately 50 kilometers in the Martian atmosphere. This extends the altitude region to well above that previously sensed.
Two spectral features have been detected toward the Kleinmann-Low infrared nebula in Orion that agree in frequency with the 7(25)-7(16) and 7(35)-7(26) rotational transitions of the ring-structured molecule ethylene oxide (CH2)2O. A feature agreeing in frequency with the 8(24)-8(45) transition probably also was detected; however, seven other transitions in (CH2)2O were not. Additional unidentified lines were detected in Ori A and Sgr B2. Searches for H(C-14)N and (C-14)O in the luminous carbon star IRC +10216 were unsuccessful. However, an unidentified line was found in IRC +10216 that was not detected in the KL nebula.