The operator which uses edge detective template matching technique is newly developed to detect linear patterns in the Landsat MSS image. This operator is applied to the Landsat image of Izu peninsula and we find that it has succeeded to detect the known linearments. Also, new type of display technique using HSI(Hue, Saturation and Intensity) color model is developed to enhance the linear pattern information, where the linear pattern direction and intensity are assigned to the Hue and the Intensity respectively.
Located in the east coast of northern Honshuu Island, Miyako Bay enjoys lots of fishery harvest, e.g., salmon, oyster, ear-shell, etc. A remote sensing project of Miyako Bay was undertaken from 1979 to 1983. During the time, 13 sets of airborne MSS data and sea truth data were collected for various seasons and tidal conditions. The data were analyzed from two points. Through regression analysis, one is for evaluation of remote sensing for water quality detection. Water temperature is effectively detected by thermal band image, but results for water color items, e.g., Secchi disk depth, chlorophyl a concentration, SS etc., were very poor. The other point is to investigate water change mechanism of the bay by using remotely sensed water temperature images and sea truth data of salinity. It is found that Miyako Bay takes several water exchange patterns depending on conditions of tide, wind and coastal current. These results are divided in two parts. This Part I surveys the project activity, and describes the results of regression analyses and water exchange mechanism in Spring time. In part 2, water exchange mechanisms in other seasons are described.
This paper is concerned with an investigation of the water exchange mechanism of the Miyako Bay by using remote sensing. As a part of Sanriku coastal zone remote sensing observation project, Miyako Bay has been observed repaetedly for the last 7 years. Nowaday we have collected 13 sets of MSS image date attached with coincidents sea truth data for different seasons and under different tidal conditions. Those data were analyzed from various view points. In the paper with the same title of part 1, the project activity, results of regression analyses between imaged data and sea truth data, and interpretation for water exchange mechanism in Spring time are described. In this part 2, water exchange mechanisms in other seasons are described. According to a specific topographical shape of the bay, Miyako Bay has several variations of water exchange mechanisms depending upon the temporal conditions of tide, wind, and coastal current. Remote sensing is very effective for the observation of surface water behavior.
The possibility of short-time forecasting of snowfall by a Mie lidar, which can measure the depolarization ratio, is discusssed. The information which can be obtained by lidar observation on snow clouds are the integrated intensity of backscattered signal, ceiling height, information of wind velocity which is obtained from the short-time variation of signal intensity and the depolarization ratio. Among these, it was found in field studies (Yonezawa) that the time variation of the ceiling height is most useful, to forecast the beginning of snowfall. By combining lidar data which was obtained in Yonezawa and in Tokyo, it was found that the ceiling height began to drop approximately 2 hours before the snow fell when the temperature on the ground was less than 0°C. This implies that snowfall can be predicted about 2 hours in advance.
This paper describes properties of TM imageries in the Suo Sea in comparison with MSS imageries. Band 2 and Band 6 imageries are found to be most useful for obtaining better informations on sea profiles, such as flow vector and temperature distribution on the water surface as expected. For both two scenes used here, Band 6 imageries reveal the vectors of oceanic water flowing into the Suo Sea through the Kanmon Strait, while Band 2 imageries are capable for drawing informations in onshore and offshore. In addition to the flow vector, Band 6 imageries seem to reveal a probable thermal contamination in onshore due to industries located around Ube and Nakatsu cities.
A framework of questions and significant points to be considered is outlined for those in developed countries who advise developing, countries on their problem solving with remote sensing. Adopting the best remote sensing approach will require problem definition and the responses to the following four questions; 1. What are the targets? 2. What level of interpretation is needed? 3. How can the targets be sensed? 4. What are the available resources and required data? In addition to these questions and responses, the following 6 points should be considered; 1. The need for a human interpreter increases as the required level of interpretation increases. 2. Spectral and computer approaches are valuable when seeking surface information, but they are often of little or no real value when. seeking subsurface information--the latter being more dependent on human interpretation. 3. As the required spatial resolution and spectral sensitivity approach the limits of the sensor, the need for applying the original remotely sensed data increases. 4. A remote sensing approach cannot produce results of some specified reliabilty and geometric accuracy unless all input data are of that reliability and accuracy. 5. The costs for data and analysis should be justified and minimized. 6. Advanced technology should be integrated and amalgamated with conventional technology. The influence of recent advances in computer technology and satellite sensors on approaches to problem solving is also discussed. It is highly recommended that the points emphasized here should be considered and the best approach should be adopted in need of developing countries.
A method to create a perspective image from Landsat TM data was descrived and the prospect was discussed. The digital terrain model (DTM) for the perspective image was a grid type. The ground elevation of it was interpolated at an interval of 40 meters for the TM data resolution from an originally produced DTM for Landsat MSS data with an interval of approximately 80 meters. Compared with a perspective image from Landsat MSS data, this newly produced one from Landsat TM data informs us the detailed feature of land. Also new color composites in the perspective image, i.e., true color one, or middle infra-red one, can be produced because of the arrival of new band images in the TM data. Color perspectives drawn in these new color composition give us a kind of different feeling. True color perspective is free for our mind. Infra-red color one is fine to recognize the vegetation on the land. To the contrary, middle infra-red one gives us good views on the skirts of moutain but in the snow covered area of the summit. Merits of the perspective image from Landsat TM data are prospectively presumed as below; 1. In the recognition of relationship between vegetation and topography. 2. In the recognition of river basin. 3. In the recognition of the topography of the land.
A Stereoscopic image viewing the Mt. Fuji and the surroundings was produced by combining Landsat TM data and the corresponding digital terrain model (DTM). The DTM consists of elevations read out at every intersections of grid divided into 150 along longitude direction of a 1 : 25, 000 topographical map and also into 120 along latitude one. Elevation corresponding to each pixel of TM data was interpolated by bi-linear method. Three cases of stereoscopic image were made; one from vertical direction, one from oblique direction, and the other from horizontal direction. Stereoscopic. image from vertical direction has different characteristics for human eyes according to the different base height ratio. In the case of Mt. Fuji viewed from an altitude of 160 km with a base length of 40 km, the terrain feature looks like natural. This case is equivalent to base height ratio of 0.25 when an object is seen from a distance of 25 cm, i.e., the distance of distinct vision. On the contrary, when it is viewed from an altitude of 50 km with the same base lengh of 40 km, the topography is emphasized. This stereo pair may be useful to detect the change of slope and also the small terrain feature. Oblique stereoscopy gives a panorama which feels more comfortable for us, probably because we are familliar with such slant vision as seen from the top of mountains. Horizontal stereoscopy gives a kind of different feeling in our terrain recognition, as it were viewed on the surface of the other planet in some space fictions, because we have no experiences to see the land surface in such clear atmosphere.