This paper deals with improvements of aperture synthesized microwave radiometer. It is known that this type of radiometer is effective to obtain higher spatial resolution than real aperture radiometer. However, two problems, that is, sufficient sidelobe reduction for practical use and pattern variation due to incident brightness temperature, are still unsolved. In this paper, a new configuration of the aperture synthesized microwave radiometer for improving the above problems is described. Experimental results using an L-band prototype radiometer are also presented.
The linear algebraic analysis of remotely sensed multispectral image is studied. The inner product is applied to the classification and the outer product to systematic treatment of mixels. The analysis uses essentially angular distribution of data with respect to the origin of feature space. The offsets which are involved in the data are estimated by the 3-dimensional histogram and subtracted from the data. Then the fact that pixels have the same angle in the feature space strongly implies that they have the same reflectances and belong to the same category. Therefore the angles of data is the most important information to classify the categories.
Reports of polarization of light by plant canopies in relation to their agronomic characteristics are scarce. In this study radiance responses of an eight band field spectroradiometer, expressed as reflectance factors, were measured using rotary polarizer and responses were related to the ratio of the legume green dry biomass to the total (legume and grass) green dry biomass (L/T) for five plots of a mixed seeding of pasture grasses (Lolium italicum, Dactylis glomerata, and others) and legumes (Trifolium repens and T. pratense) at Tochigi, Japan. Wavelength bands of the field spectroradiometer were centered at 491, 560, 662, 833, 1100, 1200, 1648 and 2210 urn. Degree of polarization (P) was calculated from the maximum and minimum of reflectance factors that were acquired by manually rotating the polarizer filter wheel from 0° to 160° at 10° intervals. The field of view at the canopy surface was 30 cm in diameter and observations were made looking toward the Sun from a zenith view angle of 45°. Each of the five plots was measured four times on May 9, 1994. During the measurement intervals the angle of incidence (the average of the view and solar zenith angles) were 32.5-33.0, 34.0-35.5, 36.0-38.5, and 39.5-43.0°. Total green dry biomass for the 5 pasture plots varied from 300 to 464 g/m2 and legume green dry biomass ranged from 12 to 220 g/m2. Observed ratios of L/T in the plots were 0.04, 0.05, 0.35, 0.46, and 0.70. Reflectance factors for the 491, 560, 662, 833, 1100, 1200 and 1648 nm bands, P for the 491, 560, 662, and 833 nm bands, and four vegetation indices-NDVI (two band difference/sum ratio for the 833 and 662 nm band reflectance factors), RVI (ratio of 833 and 662 nm reflectance factors), MSI (ratio of 1648 and 833 nm reflectance factors) and DVI (difference between 1100 and 1200 nm reflectance factors)-were statistically tested to detected differences in L/T among the pasture plots. Only NDVI, RVI, and P at 560 and 662 nm were statistically significant, and they were not significantly affected by angle of incidence. Whereas NDVI and RVI saturated at L/T greater than approximately 0.5, P at 560 and 662 nm increased almost linearly as L/T increased. In the linear regression between P and L/T, the coefficient of determination, r2 was higher for 662 nm (0.73) than for 560 nm (0.60). Among the radiometric variables tested in this study, P at 662 nm was the most promising index to predict L/T.
An assessment of the influence due to footprint changes in microwave-radiation estimation taking Antenna Pattern Correction (APC) into account has been conducted. The APC is a powerful technique for obtaining precise data of microwave-radiation, though the pattern correction is much more sensitive to the footprint changes and additive noise. Therefore it is necessary to reduce the instability that is called 'ill-posed problem in inverse problem'. In this paper, we propose an iterative algorithm for APC based on Projection Onto Convex Sets (POCS) method and show its advantages by several numerical experiments based on the parametric data of the Advanced Microwave Scanning Radiometer (AMSR) of the Advanced Earth Observing Satellite-II (ADEOS-II) which is planned to be launched in 1998. This algorithm uses a priori information to provide the convergence of the manifold including the solution iteratively. In this work, the changes in the foot-prints or the point-spread functions of AMSR along the orbit of the satellite (ADEOS-II) and the constraints in non-negative and smoothness conditions on brightness-temperature and introduced into the APC process as the a priori information. From the results of numerical experiments, the following advantages can be obtained. the ringing artefact along the temperature boundry can be reduced in comparison with those in the inversion by Fourier technique, higher spatial resolution can be obtained without large scale matrix computation, the levels of the spurious peaks in the proposed APC algorithm are reduced by 3 to 6 dB under the noisy environment in comparison with those in the direct inversion algorithm.
The Representativeness and performance of the training data are the important factors to discuss the accuracy of the supervised classification for the satellite multispectral data. In this study, a selecting method for improving the representativeness of the training data, which is called IMR Metod (selecting method for IMproving the Representativeness of training data using nPDF algorithm), was proposed. This method is based on the nPDF (n-Probability Density Functions, H. Cetin, 1991) algorithm which is very useful for multi-dimensional data transformation and reduction. The procedures of IMR Method consist of five steps as follows : STEP-1) Preparing the preliminary designated training data through the detailed ground truth. STEP-2) Classifying the data with Maximum Likelihood Classification (MLC). STEP-3) Displaying the nPDF Plot of the data distribution in each classes. STEP-4) Making the GST (Guide image for Selecting Training class) image for selecting new training classes, which are represented with different colors in accordance with nonoverlapping and overlapping areas between classes in the nPDF feature space. STEP-5) Appending new training classes and reclassifying with MLC. The effectiveness of IMR method are verified for the HRV data. The summaries of distinctive features applying IMR Method are as follows : 1) Through GST images, everyone can easily find out classes with large variance in multi-dimensional feature space and append new classes to the preliminary designated training classes. 2) In case of applying the IMR Method, the accuracy of PCC (Probability of Correct Classification) was increased to 84.9% from 80.2%. This means that the representativeness of training data was improved. 3) Furthermore, the iterative procedure for appending the training classes using IMR Method give asignificantly higher classification accuracy. 4) For improving the performance of the training data, the IMR Method (resampling method for IMproving the performance of training data using maximum likelihood method and iso-data algorithm) was already proposed by ourselves in 1993. In case of combining IMP Method with IMR Method, it was confirmed the accuracy of PCC was increased to 92.6% from 80.2%, that is to say, the best classfication accuracy was achieved. In conclusions, this combined classifers proved to be a indispensable and practical technique for the improvement of the multispectral classification accuracy.
Tropical Rainfall Measuring Mission (TRMM) is a joint space program between USA and Japan to measure rainfall of tropics where about 60% of global rainfall is concentrated. TRMM is the first space mission dedicated to measurements of tropical rainfall with the first precipitation radar in space. Communications Research Laboratory has made the system study of the precipitation radar for the TRMM in the joint feasibility study of TRMM between USA and Japan and moreover in the follow-on studies. The basic system parameters of TRMM precipitation radar, such as frequency, antenna beam width, transmitting peak power, scan angle, horizontal and vertical resolutions, pulse repetition frequencies and so on are determined in order to satisfy the mission requirement shown by NASA. The precipitation radar system must adopt the high speed electric scanning in order to observe a raining area without any gaps between scanning lines that are perpendicular to the moving direction of the satellite. The frequency agility technique, which uses dual frequency separated 6 MHz, is adopted to attain the required independent sample number because the dwell time of the antenna beam in one angle-bin direction is limited. In the trade-off studies, three major competitive items were discussed for the selection of a preferred candidate for TRMM radar. These competitive items are (1) a passive array radar with a TWTA (Traveling Wave Tube Amplifier) versus active array radar with SSPAs (Solid State Power Amplifiers), (2) a pulse compression radar versus conventional type radar, and (3) antenna type (planar array antenna versus cylindrical parabolic antenna). Trade-off studies of these items show that the non-pulse compression active array radar with planar array antenna is considered as the most suitable candidate for the TRMM precipitation radar at 13.8 GHz.