Taking the fluorescence spectra is a useful method for the study of green leaf, because the fluorescence spectra give us the information about microstructure of photosystems of the leaf. The spectra of fluorescence induced by UV Ar-laser lights had a peak around 450 nm (F 450), a shoulder around 530 nm (F 530) and peaks of chlorophyll at 687 nm (F 687) and 741 nm (F 741). In order to examine the transverse distribution of the blue-green (450 nm, 530 nm) and the chlorophyll fluorescence (687 nm, 741 nm) within the leaf tissue induced by UV and visible laser lights projected on the intact leaves of camellia tree, a micro-fluorescence imaging (MFI) system was devised using a microscope, a CCD camera with an image intensifier, a microcomputer and laser light sources (Ar laser and He-Ne laser).
The fluorescence of 450 nm and 530 nm were observed at the epidermal cell, and the chlorophyll fluorescence (687 nm and 741 nm) at the palisade cell. Because the epidermal cell does not contain chlorophyll, fluorescence (450 nm and 530 nm) do not relate to the photosynthetic reaction directly. It is considered that these results are due to compounds in the leaves accumulated according to physiological characteristics of each plant species. We took the fluorescence spectra of young and old leaves of camellia tree by LIF system which was composed of Ar-lasers (351-364 nm, 488 nm and 514.5 nm) and a He-Ne laser (632.8 nm) for excitation, and an optical multichannel spectrometer. From the result of these, the relations between LIF spectra of plant leaves and transverse distribution of blue-green and chlorophyll fluorescence within leaf tissue were described. We conclude that imaging of transverse distribution of fluorescence in the intact plant leaves is a powerful method for the study of LIF spectra to examine activity and ecological status of plants by remote-sensing.

View full abstract

An investigation was tried to measure the affect of high concentration air pollution on boardleaf tree by means of spectral reflectance. The relationship between spectral reflectance and the amount of sulfur dioxide (SO₂) deposited in leaves of the poplar tree were measured at the neigborhood of Tonghua city in China. An function was made by discriminant analysis for the relationship between the groups of SO₂, deposited in leaves and vegetation indexes. The conclusions of this study are as follow.
The concentration of air pollution and amount of SO₂, deposited in leaves of poplar decrease in accordance with the distance from atmosphere pollution source, and the vegetation indexes (RVI and NDVI) are in inverse proportion to the amount of SO₂, deposited in leaves. In proportion to the amount of SO₂, deposited on leaves of poplar, the reflectance at Red wavelength (400-700 nm) of the leaves becomes high, and the reflectance at Near-infraned wavelength (750-950 nm) of the leaves becomes low. The degree of amount of SO₂, deposited in leaves can be estimated by the function of vegetation indexes of (RVI and NDVI), the formula is expressed as Y=0.595(RVI)+0.005(NDVI).

View full abstract

An X-band real aperture side looking airborne radar (SLAR) was developed to study the performance of the SLAR to detect the ocean oil slicks. This system has two features intended to improve the performance of ocean oil slick observation ; a modified cosecant-squared antenna pattern and an antenna steering function. In order to achieve quantitative radar observation, external calibration of the SLAR was performed using corner reflectors and sea surface as a reference target. Ocean oil slick observation experiment was also performed using oleyl alcohol to generate slicks. Through this experiment, high sensitivity of the SLAR to detect the oil slick has been confirmed. It is also demonstrated that the antenna steering capability of the SLAR is useful to optimize the SLAR swath width under variable weather conditions and flight altitude restrictions. SLAR observation parameters suitable for the ocean oil slicks are discussed based upon the experiment results and theoretical findings.

View full abstract

In this paper, we propose four kinds of landcover classification models using neural networks which are driven by co-occurrence matrices. The first and the second models classify each band image using a neural network at the first stage, then perform a final decision based on the first stage results. At the second stage, an arithmetic decision algorithm is used in the first model, and a second neural network is used in the second model. The third model has individual input and hidden layers for each band. However, the output layer of the third model collects all hidden layer outputs. The fourth model is a single stage classifier, and co-occurrence matrices of all bands are inputted into an input layer at the same time. In order to evaluate the performance of these proposed models, landcover classifications using these models and three kinds of conventional classifiers were conducted for Landsat TM and SPOT HRV data. The conventional methods used are a maximum likelihood classifier and two kinds of neural network classifier which use a single pixel and 3X3 pixels as input data. As a result, the best performance was shown by the fourth model, which showed 3% to 21% and 9% to 20% higher accuracy than the maximum likelihood classifier and the conventional neural network classifiers, respectively.

View full abstract

It is generally desired to estimate wind aloft by the use of satellite successive images of clouds and this estimate is called "cloud motion wind" (CMW). The CMW, however, has been known sometimes very erroneous. This paper aims to estimate the causes of error statistically and analytically, and to improve the CMW.
It is found that most of the wrong CMWs aloft are associated with the jet stream, the local steep slope in the tropopause and the frontal layer frequently appearing just below the jet stream core.It should be noted that, related to the meteorological features stated above, slight failure in assigning the cloud height would give larger error to the CMW.
Based on those findings, the original CMW data are applied a kind of quality control and modified with the help of winds observed by rawinsonde, if available. With this method, the errors are found to be about 8m/s, on the average, though they are about 12m/s without this.

View full abstract

Classification of remote sensing multispectral data empolying the angular distribution of the data in the feature space is proposed. Several types of the classifiers are tested. The best one also involves radial coordinate. Their reliability stems from the assumption that same ratios of reflectances of remotely sensed data strongly imply that reflectances are actually same. Data whose ratios of reflectances are same are on the same line. They belong to the same category, if the assumption is definite. The line passes near the origin of the feature space, however, not exactly on the origin. It is due to the offsets of the data, which must be carefully handled in our method.

View full abstract