This article summarizes the present state of the application of synthetic aperture radar (SAR) to agriculture. Among the various applications of spaceborne SARs, agriculture is one of the important fields in which the characteristics of SAR, including all-weather and day-and-night imaging capabilities, can be fully utilized on a global scale with high spatial resolution. The main objectives to be achieved by SAR in the field of agriculture are to estimate the agricultural crop list and production, land cover and use, and soil moisture, the source of drought, floods, and above all, crop growth. In the present review, a brief summary of the characteristics of SAR and its parameters associated with agricultural applications is given, and the methodologies for estimating growth and classifying crops are then described. The techniques of soil moisture measurement are then described with examples. Although this review article deals with the up-dated theories and examples, some early studies are also included, as they are the basis of the present technology.
With the advent of new remote-sensing platforms and sensors such as drones, Synthetic Aperture Radar (SAR), high-frequency observation satellites, etc., remote sensing technology is now being used to elucidate or resolve many agriculture-related issues from global food security to precision farming. This article looks back on the trajectory of the remote sensing studies on agriculture in Japan and reviews the results of studies published mainly by researchers at agricultural research institutes in four time periods since the late 1970 s, which we refer to as the “dawn and infancy period”, “development period”, “establishment period”, and “present”.
The estimation of paddy rice field areas is an important application of satellite remote sensing in agricultural research. Polarimetric data are useful in this regard and can be obtained from the PALSAR-2 L-band synthetic aperture radar (SAR) system aboard the ALOS-2 satellite. In this study, PALSAR-2 data acquired in the full polarimetric mode during the maturation season of rice were analyzed in order to classify agricultural areas according to their use for paddy rice and other crops, such as soybeans. Eigenvalue-eigenvector and four-component decompositions were used to classify the PALSAR-2 data and discriminate between different agricultural parcels. Vector data for agricultural land-use areas were overlaid on the analyzed images, and the mean value for each agricultural parcel was computed. Landsat 8 OLI images obtained during flooding and heading times of paddy rice fields were also analyzed in order to extract paddy rice agricultural parcels. A quantitative comparison of the classification results made it possible to determine the most efficient decomposition components for discriminating paddy rice fields from fields of other crops: alpha angle, double bounce scattering component ratio, and surface scattering component ratio. Mature paddy rice plants cause linear dipole scattering, which allows differentiation between paddy rice and other crops. The spatial distribution maps of these parameters’ threshold values corresponded well with the Landsat 8 OLI analysis results. For example, a comparison of the alpha angle classification with the Landsat data resulted in differences of 1.8 % and 1.6 % in field number and area, respectively. One advantage of full polarimetric data analysis is the usefulness of one-time observation data for paddy field extraction. This study’s results prove the usefulness of PALSAR-2 full polarimetric data in discriminating full-grown paddy rice fields from fields of soybeans and other crops.
Satellite-based remote sensing would provide useful geo-information for the information-based agricultural management (Smart Agriculture). This paper reviews the background and information needs for smart agriculture as well as the necessary sensor specifications and algorithms for application of high-resolution satellites to smart agriculture. The generalized spectral index approach based on hyperspectral data can be used to determine optimal algorithms for prediction of plant and soil variables. This approach would be useful to enhance the applicability of optical satellite sensors that have different waveband specifications. The efficient use of high-resolution satellite sensors would strongly support the diagnostics and decision making in smart agriculture at regional scales.
Drone-based remote sensing has a great potential for spatial diagnostics of crops and soils in the information-based agricultural management (Smart Agriculture). This paper reviews the background and advanced applications of drone-based remote sensing to smart agriculture. The necessary performance of optical and thermal sensing systems, data processing techniques, and algorithms for assessment of crop and soil variables are discussed based on case studies. Drone-based remote sensing would be useful for timely and efficient monitoring of crops, soils, weeds, diseases, etc., mainly in farms on the scale of ∼100 ha. Sophisticated systems and workflow have to be established to provide not only useful on-demand information but also sufficient cost-benefits.
In this study, we compared pixel-based image analysis and object-based image analysis (OBIA) as methods of land cover classification of urban areas, using high resolution digital aerial photography. The study area was Setagaya Ward, Tokyo, Japan, and we carried out supervised classification using aerial photographs with 25-cm spatial resolution, and with both visible bands and a near infrared band. The overall accuracy of the object-based classification was approximately 6 to 20 percentage points higher than that of the pixel-based classification. Both methods tended to over-classify water areas and bare land, specifically, with shadows of buildings and roads in impervious areas tending to be misclassified as water areas and as bare land, respectively. The tendency of over-classification was remarkable in the pixel-based classification, and most of the classified area was a minute area of 1 pixel to several pixels. To evaluate such minute areas (salt-and-pepper effect) in the classification, we calculated the join-count statistics, a kind of spatial autocorrelation index. The tendency of grouping of the object-based classification was stronger than of the pixel-based classification, indicating that the minute areas in the object-based classification were fewer than those in the pixel-based classification. We also calculated the green coverage ratio based on the present classification results, and compared it to that obtained by the municipality of Setagaya Ward. The green coverage ratio from the object-based classification was between the two values by the municipality, whereas the value from the pixel-based classification was smaller than both values. In conclusion, the object-based method is applicable to the land cover classification and extraction of vegetation using high-resolution digital aerial photography in urban areas.
Yangon, the former capital of Myanmar, is the major economic areas of the country. Also, the urban areas have significantly increased. However, Yangon has problems with disasters such as flood and earthquake. To support disaster risk management in Yangon, Myanmar, the estimation of urban expansion is required to understand the mechanism of urban expansion and predict urban areas in the future. This research proposed a methodology to develop urban expansion modeling based the dynamic statistical model using Landsat Time-Series and GeoEye Images. Multispectral Landsat images from 1978 to 2009 were classified to provide land cover change with a long period. By observing land cover from the past to the present, the class translation matrix was obtained. Stereo GeoEye Images in 2013 were employed to extract the heights of buildings. By using the heights of buildings, the multi-centers of urban areas cloud be detected. The urban expansion modeling based on the dynamic statistical model was defined to refer to three factors; (1) the distances from the multi-centers of the urban areas, (2) the distances from the roads, and (3) the class translation. The estimation of urban expansion was formulated in term of the dynamic statistical model by using the maximum likelihood estimator. The relevant equations to estimate urban expansion are expressed in this research. The prediction of urban expansion was defined by the combination of the estimation of urban expansion and the estimated parameters in the future. In the experiments, the results indicated that our model of urban expansion estimated urban growth in both estimation and prediction steps with efficiency.