梶原 康司

We conducted a high-resolution aeromagnetic survey using an autonomously driven uncrewed helicopter that flew as low as several tens of meters above the ground along precise flight tracks with 1 m accuracy. The geomagnetic total intensity was measured by a total intensity magnetometer suspended beneath the helicopter at a ~ 50 m or less flight spacing over the entire caldera of Mt. Mihara, located on Izu-Oshima Island, Japan. From the observed geomagnetic data, we estimated high-resolution subsurface magnetization intensity. A high average magnetization intensity of 13.5 A/m was obtained for the entire caldera. The distribution of the magnetization intensity was not only consistent with the results of conventional airborne surveys, but it also had a high spatial resolution of less than 100 m. Highly magnetized areas were observed along the NW–SE lines that intersected the summit pit crater, Crater A, which is consistent with the principal stress direction of Izu-Oshima Island. These highly magnetized areas might be solidified magma that did not reach the surface during past eruptions. A large and deep-rooted weakly magnetized area was found just outside of the NE side of the central cone, which corresponds to the location of Fissure B, and the conduit must have been demagnetized at the previous event. Other weakly magnetized areas were also observed at the N, E, and SW sides around the pit crater. These regions correspond to the location of fumaroles in the crater. The high-resolution subsurface magnetization imaged by the autonomous uncrewed helicopter will be helpful for the mitigation of future eruption damage by enabling the assessment of potential fissure eruption areas.

Abstract. This paper examines algorithms for estimating terrestrial albedo from the products of the Global Change Observation Mission – Climate (GCOM-C) / Second-generation Global Imager (SGLI), which was launched in December 2017 by the Japan Aerospace Exploration Agency. We selected two algorithms: one based on a bidirectional reflectance distribution function (BRDF) model and one based on multi-regression models. The former determines kernel-driven BRDF model parameters from multiple sets of reflectance and estimates the land surface albedo from those parameters. The latter estimates the land surface albedo from a single set of reflectance with multi-regression models. The multi-regression models are derived for an arbitrary geometry from datasets of simulated albedo and multi-angular reflectance. In experiments using in situ multi-temporal data for barren land, deciduous broadleaf forests, and paddy fields, the albedos estimated by the BRDF-based and multi-regression-based algorithms achieve reasonable root-mean-square errors. However, the latter algorithm requires information about the land cover of the pixel of interest, and the variance of its estimated albedo is sensitive to the observation geometry. We therefore conclude that the BRDF-based algorithm is more robust and can be applied to SGLI operational albedo products for various applications, including climate-change research.

In situ accurate data sets of leaf area index (LAI), above-ground biomass (AGB), and fraction of absorbed photosynthetically active radiation (fAPAR) are indispensable to validate and improve ecological products obtained from satellites. In situ data for satellite validation must be created not from a single-point data but from areal data (such as multiple-points data) representing a satellite footprint. Using multiple-points data, the error of in situ data can be calculated statistically. The quantification of the error in the in situ data enables us to evaluate the discrepancy between the satellites' products and the in situ data as the error in the in situ data and the estimation error in the products separately. Besides, the accuracy of the in situ data is required to be much higher than the accuracy of the satellite products which was officially set. To obtain such in situ data, we have established observation sites for typical land cover types in East Asia, from temperate to cool ecosystems: deciduous needle-leaved forest (DNF), evergreen needle-leaved forest (ENF), deciduous broad-leaved forest (DBF), and grassland (GL). We conducted the observations in 500 m x 500 m areas, which is the footprint scale of the Global Change Observation Mission-Climate satellite. In this paper, the data of LAI, AGB, and fAPAR observed at DNF, DBF, and GL (i.e., except at ENF) are reported. These data are useful even for the validation of other satellite products, especially with higher spatial resolution. Also, the long-term tree census data from 2005 to 2018 at DNF are reported. The complete data set for this abstract published in the Data Paper section of the journal is available in electronic format in MetaCat in JaLTER at .

To validate and to improve ecological products obtained from satellites, such as a leaf area index (LAI), above-ground biomass (AGB), and a fraction of photosynthetically active radiation (fAPAR), in-situ accurate data are indispensable. They must be not a single point-data but an areal data representing the satellite footprint. Their accuracy needs to be much higher than the required accuracy for the satellite products. The quantitative assessment of their error is necessary for evaluating the satellite products' error from the discrepancy between the satellite products and the in-situ data. However, such data had not been available. In particular, there had been few data of LAI in a sparse evergreen needle-leaved forest, because of difficulty of accuracy control of in-situ observation in such a forest. To overcome the difficulty and to obtain the representative LAI, we made an allometric equation to estimate the leaf mass of Picea glehnii in northern Hokkaido. We report the allometric equations of leaf mass and AGB of P. glehnii, its leaf mass per area (LMA), its leaf life span, its leaf distribution, its crown shapes, its wood specific gravity, and tree locations. We also report LAI, AGB, and fAPAR within the 500 m x 500 m area, which is the footprint scale of the Global Change Observation Mission-Climate satellite, in a pure and sparse forest of P. glehnii in northern Hokkaido. These precise data are useful for validation of other satellite data, especially with higher spatial resolution, and forest structure modeling.

<p>In this paper, we present an algorithm for estimating terrestrial albedo for the product of Global Change Observation Mission-Climate (GCOM-C)/Second generation Global Imager (SGLI), that was launched in Dec. 2017 by Japan Aerospace Exploration Agency (JAXA), Japan. The algorithm is composed of spectral albedo estimation, narrowband-to-broadband albedo conversion and multi-regression model estimation so that only a single-day reflectance observation is available. In estimating spectral albedo, we derives coefficients of kernel-driven bidirectional reflectance distribution function (BRDF) model. The experiments by using in-situ data of bare soil and deciduous broadleaf forests show that the proposed method have potential to estimating albedos with acceptable accuracy of the root mean square of errors (RMSE) of 2.2×10^-2 and 4.3×10^-2.</p>

To validate terrestrial ecological products of the Global Change Observation Mission-Climate (GCOM-C) satellite, a large-scale ecological observation project "JAXA Super Sites 500" was initiated. The project's purpose is to obtain the representative leaf area index (LAI), above-ground biomass (AGB), and fraction of absorbed PAR (fAPAR) in the satellite footprint scale. This study aimed to determine the appropriate observation methods in the satellite scale for the target land-cover types such as a deciduous broad-leaved forest, evergreen/deciduous needle-leaved forests, and a grassland. As a result of the comparative observations, a combination method using litter-traps and LAI-2200 was adopted as the LAI observation in a forest, a tree census method was adopted as the AGB observation in a forest, and a clipping method was adopted as the AGB and LAI observations in a grassland. LAI, AGB, and fAPAR were observed and were used to validate the GCOM-C terrestrial ecological products.

Japan Aerospace Exploration Agency (JAXA) had launched new Earth observation satellite GCOM-C in the end of 2017. The core sensor of GCOM-C, Second Generation Global Imager (SGLI) has a set of along track slant viewing Visible and Near Infrared Radiometer (VNR). These multi-angular views aim to detect the structural information from vegetation canopy, especially forest canopy, for estimating productivity of the vegetation. SGLI Land science team has been developing the algorithm for Above Ground Biomass, Vegetation Index, Shadow Index, etc.We developed the ground observation method using Unmanned Aerial Vehicle (UAV) in order to contribute the algorithm development and its validation. Mainly, multiangular spectral observation method and simple BRF model have been developed for estimating slant view response of forest canopy. The BRDF model developed by using multi-angular measurement has been able to obtain structural information from canopy. In addition, we have conducted some observation campaigns on typical forest in Japan in collaboration with other science team experienced with vegetation phenology and carbon flux measurement. Primary result, Above-ground biomass of these observations are also be demonstrated.

We report long-term continuous phenological and sky images taken by time-lapse cameras through the Phenological Eyes Network (http://www.pheno-eye.org. Accessed 29 May 2018) in various ecosystems from the Arctic to the tropics. Phenological images are useful in recording the year-to-year variability in the timing of flowering, leaf-flush, leaf-coloring, and leaf-fall and detecting the characteristics of phenological patterns and timing sensitivity among species and ecosystems. They can also help interpret variations in carbon, water, and heat cycling in terrestrial ecosystems, and be used to obtain ground-truth data for the validation of satellite-observed products. Sky images are useful in continuously recording atmospheric conditions and obtaining ground-truth data for the validation of cloud contamination and atmospheric noise present in satellite remote-sensing data. We have taken sky, forest canopy, forest floor, and shoot images of a range of tree species and landscapes, using time-lapse cameras installed on forest floors, towers, and rooftops. In total, 84 time-lapse cameras at 29 sites have taken 8 million images since 1999. Our images provide (1) long-term, continuous detailed records of plant phenology that are more quantitative than in situ visual phenological observations of index trees; (2) basic information to explain the responsiveness, vulnerability, and resilience of ecosystem canopies and their functions and services to changes in climate; and (3) ground-truthing for the validation of satellite remote-sensing observations.

Forest disturbance by heavy snow seriously affects ecosystem functions and provision of ecosystem services. To evaluate the spatial distribution of this disturbance over large areas, it is necessary to develop a flexible, inexpensive, and generalizable method based on remote sensing. Here, we examined the ability of an unmanned drone to detect the disturbance caused by heavy snow in a Japanese cedar (Cryptomeria japonica) forest, which is a typical landscape species in Japan's mountainous areas. We obtained aerial photographs in late October 2016 using the drone in a research plot where many individuals were damaged by moist, heavy snow in mid-December 2014. The forest disturbance rate was estimated by visually inspecting the structure from motion (SfM) point clouds generated from the drone's aerial photographs. We detected 90 to 96% of healthy individuals, but many tilted trees and trees with broken stems but an intact canopy were misidentified as healthy individuals. The estimated forest disturbance rate (33%) obtained from the SfM point clouds coincided well with the actual forest disturbance rate (35%) obtained from tree surveys. Consequently, this approach can potentially be used to detect narrow and patchy disturbances in Japanese cedar forest, although further observations at multiple points will be required to develop the accuracy of this approach.

We constructed a bi-directional reflectance distribution function (BRDF) observation system on a 25-m-high tower surrounded by mature larch forests at the foot of Mt. Yatsugatake, central Japan. The BRDF feature (strong/weak reflections in a backward/forward scattering areas) is related closely to the 3D structure and biophysical parameters of forests. However, because of the difficulty of observations, observational BRDF data of mature forests are scarce, except for those of Boreal Ecosystem -Atmosphere Study (BOREAS). We made BRDF observations 10 times during 2007 to ascertain BRDF characteristics in each season. For comparison to these BRDF data, a semi-empirical BRDF model (Ross-thick-Li sparse kernel model) was introduced. This model can reproduce the BRDF data fairly well under identical solar conditions in each season. Then, seasonal changes of the vegetation structure and possible relations with BRDF were discussed. They are summarized as follows: (1) The forest BRDF changes distinctly and seasonally depending on the phenology. (2) The forest BRDF is influenced primarily by clarity of the geometrical profile on the forest surface, which is affected by shadow. However, the forest BRDF might be influenced by other factors such as the physiological characteristics of leaves and slight change of vegetation structure such as forest floor conditions and leaf angles during summer. (3) Features of forest BRDF must therefore appear during summer when the crown shape is distinctive. Furthermore, BRDF features appear distinctly as seasons progress, e.g., by autumn leaves and defoliation, when the unevenness of forests increases. These findings can promote monitoring of vegetation using multi-angular satellite observations.

Japan Aerospace Exploration Agency (JAXA) had launched new Earth observation satellite GCOM-C in the end of 2017. The core sensor of GCOM-C, Second Generation Global Imager (SGLI) has a set of along track slant viewing Visible and Near Infrared Radiometer (VNR). These multi-angular views aim to detect the structural information from vegetation canopy, especially forest canopy, for estimating productivity of the vegetation. SGLI Land science team has been developing the algorithm for Above Ground Biomass, Vegetation Index, Shadow Index, etc.We developed the ground observation method using Unmanned Aerial Vehicle (UAV) in order to contribute the algorithm development and its validation. Mainly, multiangular spectral observation method and simple BRF model have been developed for estimating slant view response of forest canopy. The BRF model developed by using multiangular measurement has been able to obtain structural information from canopy. In addition, we have conducted some observation campaigns on typical forest in Japan in collaboration with other science team experienced with vegetation phenology and carbon flux measurement. Primary results of these observations are also be demonstrated.

For monitoring of global environmental change, the Japan Aerospace Exploration Agency (JAXA) has made a new plan of Global Change Observation Mission (GCOM). SGLI (Second Generation GLI) onboard GCOM-C (Climate) satellite, which is one of this mission, provides an optical sensor from Near-UV to TIR. Characteristic specifications of SGLI are as follows; 1) 250m resolutions over land and area along the shore, 2) Three directional polarization observation (red and NIR), and 3) 500m resolutions temperature over land and area along shore. The final check of the sensor system has been completed in 2016. The final check of the entire satellite system will be completed in the first half of 2017. The launch of GCOM - C is scheduled for the second half of 2017. These characteristics are useful in many fields of social benefits. In addition, 51 products will be made by mainly 35 principal investigators. There are 14 land products of SGLI (standard products are 9, and research products are 5). Already the Land team completed most of the algorithm and have created a product validation plan. Currently land team are improving the performance of the algorithm and preparing for verification. These activities are reported in this paper.

Since fallen trees are a key factor in biodiversity and biogeochemical cycling, information about their spatial distribution is of use in determining species distribution and nutrient and carbon cycling in forest ecosystems. Ground-based surveys are both time consuming and labour intensive. Remote-sensing technology can reduce these costs. Here, we used high-spatial-resolution aerial photographs (0.5-1.0 cm per pixel) taken from an unmanned aerial vehicle (UAV) to survey fallen trees in a deciduous broadleaved forest in eastern Japan. In nine sub-plots we found a total of 44 fallen trees by ground survey. From the aerial photographs, we identified 80% to 90% of fallen trees that were >30 cm in diameter or >10 m in length, but missed many that were narrower or shorter. This failure may be due to the similarity of fallen trees to trunks and branches of standing trees or masking by standing trees. Views of the same point from different angles may improve the detection rate because they would provide more opportunity to detect fallen trees hidden by standing trees. Our results suggest that UAV surveys will make it possible to monitor the spatial and temporal variations in forest structure and function at lower cost.

近年，世界各国の水工・水文分野において，気象予測データを用いて，河川流量・旱魃・穀物収量などの面的な農業・水資源変動予測が実施されている．これらの予測には，近未来の気象予測精度向上に加え，陸面初期値の精度向上が重要である．我々の研究グループでは，環太平洋域における農業水資源変動予測を目的に，陸面再解析・速報解析システム開発に取り組んだ．本稿では，環太平洋域を対象としたシステム開発の一環として行った，日本域における高解像度陸面再解析について述べる．解析の1つの特色は，全日本域を km解像度で計算を実行する点である．検証データが比較的豊富にある日本を対象とし，モデル解析値の検証や，感度実験による高解像度化・気象強制力差し替えの効果の調査を目的としている．

Japan Aerospace Exploration Agency (JAXA) is going to launch new Earth observation satellite GCOM-C1 in 2016. The core sensor of GCOM-C1, Second Generation Global Imager (SGLI) has a set of along track slant viewing Visible and Near Infrared Radiometer (VNR). These multi-angular views aim to detect the structural information from vegetation canopy, especially forest canopy, for estimating productivity of the vegetation. SGLI Land science team has been developing the algorithm for above ground biomass, canopy roughness index, shadow index, etc.In this paper, we introduce the ground observation method developed by using Unmanned Aerial Vehicle (UAV) in order to contribute the algorithm development and its validation. Mainly, multi-angular spectral observation method and simple BRF model have been developed for estimating slant view response of forest canopy. The BRF model developed by using multi-angular measurement has been able to obtain structural information from canopy. In addition, we have conducted some observation campaigns on typical forest in Japan in collaboration with other science team experienced with vegetation phenology and carbon flux measurement. Primary results of these observations are also be demonstrated.

Various remote sensing technologies are utilized to monitor tropical forest for REDD project. In developing countries, the accuracy of field measurement does not satisfy the quality required for the validation of satellite remotely sensed data. Therefore, we introduce the most portable laser sensor, SICK LMS511, to measure trees in the field of tropical forest and the semi-automatic process to derive tree parameters from the 3D point cloud is established to provide accurate forest inventory data. From the accuracy assessment of the portable laser system provided 3.61 cm of root mean square error (RMSE) of stem diameter and 0.50 m of RMSE of tree height measurement. The portable terrestrial laser sensor can be the standard technique to create forest inventory data for the field validation of satellite remote sensing.

The canopy shadow fraction (CSF) is composed of the fractional area covered by shadowed tree crowns and shadowed backgrounds for a given illumination and view geometry. Since the CSF is related to the canopy biological and structural features, an accurate estimation of the CSF is expected for better understanding of the canopy characteristics. This study explores an algorithm for an automated extraction of the CSF using near-surface remote sensing method. The high-spatial resolution true-color images over different forested canopies were acquired using an unmanned helicopter. For each site, the images of the same target canopy from multiple view zenith angles (VZA) were taken at the principal plane. The digital images were processed to extract the CSF using the color vegetation indices (CVI) combined with an image threshold algorithm. The CSF was measured based on visual interpretation of the grayscale images. For an automated extraction of the CSF, different CVI related to CSF were assessed with Otsu threshold algorithm. A new CVI called the blue deficient index (BDI) was proposed as an indicator of the CSF exploiting the canopy spectral properties. The performance of each automated extraction method was evaluated with comparison to measured CSF. Among the methods assessed under the study, the CSF estimated by the BDI with the Otsu algorithm was found to be most closely related to the measured CSF. After a successful extraction of the CSF, the effect of VZA on CSF was analyzed. The substantial variation of the CSF with respect to the VZA in the principal plane was confirmed for a given solar position.

Near-surface bi-directional reflectance and high-spatial resolution true-color imagery of several forested canopies were acquired using an unmanned helicopter. The observed reflectance from multiple view-zenith angles were simulated with a kernel-driven bidirectional reflectance model, and the BRDF parameters were retrieved. Based on the retrieved BRDF parameters, kernel-derived multi-angular vegetation indices (KMVIs) were computed. The potential of KMVI for prediction of canopy structural parameters such as canopy fraction and canopy volume was assessed. The performance of each KMVI was tested by comparison to field measured canopy fraction and canopy volume. For the prediction of canopy fraction, the KMVI that included the nadir-based NDVI performed better than other KMVI emphasizing the importance of nadir observation for remote estimation of the canopy fraction. The Nadir BRDF-adjusted NDVI was found to be superior for the prediction of canopy fraction, which could explain 77% variation of the canopy fraction. However, none of the existing KMVI predicted the canopy volume better than Nadir BRDF-adjusted NDVI and Nadir-view NDVI. The Canopy structural index (CSI) was proposed with the combination of normalized difference between dark-spot near infrared reflectance and hot-spot red reflectance. The CSI could establish an improved relationship with the canopy volume over Nadir BRDF-adjusted NDVI and Nadir-view NDVI, explaining 72% variation in canopy volume. In addition, MODIS based KMVI were evaluated for the prediction of canopy fraction and canopy volume. MODIS based KMVI also showed similar results to the helicopter based KMVI. The promising results shown by the CSI suggest that it could be an appropriate candidate for remote estimation of three-dimensional canopy structure. (C) 2013 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS) Published by Elsevier B.V. All rights reserved.

For monitoring of global environmental change, the Japan Aerospace Exploration Agency (JAXA) has made a new plan of Global Change Observation Mission (GCOM). SGLI (Second Generation GLI) onboard GCOM-C (Climate) satellite, which is one of this mission, provides an optical sensor from Near-UV to TIR. Characteristic specifications of SGLI are as follows; 1) 250m resolutions over land and area along the shore, 2) Three directional polarization observation (red and NIR), and 3) 500m resolutions temperature over land and area along shore. SGLI has a set of along track slant viewing Visible and Near Infrared Radiometer (VNR). These multi-angular views aim to detect the structural information from vegetation canopy, especially forest canopy, for estimating productivity of the vegetation. SGLI Land science team has been developing the algorithm for above ground biomass, canopy roughness index, shadow index, etc.In this paper, we introduce SGLI characteristics and the ground observation method developed by using Unmanned Aerial Vehicle (UAV) in order to contribute the algorithm development and its validation. Mainly, multi-angular spectral observation method and simple BRF model have been developed for estimating slant view response of forest canopy. The BRF model developed by using multi-angular measurement has been able to obtain structural information from vegetation canopy. In addition, we have conducted some observation campaigns on typical forest in Japan in collaboration with other science team experienced with vegetation phenology and carbon flux measurement. Primary results of these observations are also be demonstrated.

The early stage of the water stressed forest shows the higher temperature before the spectral reflectance change. To detect the water stressed forest, the satellite detected surface temperature is utilized. The day and night surface temperature difference is the key factor of the detection, in the case of non-stressed forest the daytime surface temperature suppress the latent heat increase and the nighttime surface temperature is almost same as the air temperature at the surface, so that the water stress makes the daytime temperature increases. The day and night surface temperature difference is primary affected by the forest water stress level. To remove the another effect to the temperature difference such as the nighttime low air temperature in autumn, the modified day and night surface temperature difference is defined for the forest water stress detection index. Using the day night surface temperature product from MODIS and the latent heat flux dataset acquired at some sites of the AMERIFLUX, The water stressed forest is identified using the proposed index. Also the numerical simulation for the sensitivity analysis of the proposed index is made and the effectiveness of the index is clarified.. © 2012 SPIE.

It is very important to watch the spatial distribution of vegetation biomass and changes in biomass over time, representing invaluable information to improve present assessments and future projections of the terrestrial carbon cycle. A space lidar is well known as a powerful remote sensing technology for measuring the canopy height accurately. This paper describes the ISS(International Space Station)-JEM(Japanese Experimental Module)-EF(Exposed Facility) borne vegetation lidar using a two dimensional array detector in order to reduce the root mean square error (RMSE) of tree height due to sloped surface.

Japan Aerospace Exploration Agency (JAXA) is going to launch new Earth observation satellite GCOM-C1 in near future. The core sensor of GCOM-C1, Second Generation Global Imager (SGLI) has a set of along track slant viewing Visible and Near Infrared Radiometer (VNR). These multi-angular views aim to detect the structural information from vegetation canopy, especially forest canopy, for estimating productivity of the vegetation. SGLI Land science team has been developing the algorithm for above ground biomass, canopy roughness index, shadow index, etc.In this paper, we introduce the ground observation method developed by using Unmanned Aerial Vehicle (UAV) in order to contribute the algorithm development and its validation. Mainly, multi-angular spectral observation method and simple BRF model have been developed for estimating slant view response of forest canopy. The BRF model developed by using multi-angular measurement has been able to obtain structural information from vegetation canopy. In addition, we have conducted some observation campaigns on typical forest in Japan in collaboration with other science team experienced with vegetation phenology and carbon flux measurement. Primary results of these observations are also be demonstrated.

For monitoring of global environmental change, the Japan Aerospace Exploration Agency (JAXA) has made a new plan of Global Change Observation Mission (GCOM). SGLI (Second Generation GLI) onboard GCOM-C (Climate) satellite, which is one of this mission, provides an optical sensor from Near-UV to TIR. Characteristic specifications of SGLI are as follows; 1) 250m resolutions over land and area along the shore, 2) Three directional polarization observation (red and NIR), and 3) 500m resolutions temperature over land and area along shore. These characteristics are useful in many fields of social benefits. In addition, 51 products will be made by mainly 35 principal investigators. We introduce the overview of GCOM-C1/SGLI science.

The results of aeromagnetic observations at Izu-Oshima, Japan using an unmanned autonomous helicopter are reported. A practical observation system was assembled, adopting a bird-type magnetometer installation, and dense observations of the northern half of the caldera area including the central cone were made from a very low altitude. In the detailed magnetization intensity mapping deduced from the collected data, low magnetization intensity at the vent and three rows of high magnetization intensity on the caldera floor were found. The former is interpreted as the presence of high-temperature materials, such as magma or hot rock, or vacant space in the conduit. The latter is considered to be due to solidified dykes. Low magnetization intensity suggesting a magma body (or subsidiary magma chamber) was not detected below the caldera floor. The observation results confirmed that aeromagnetic observation using an unmanned autonomous helicopter had sufficient performance for volcanic observations, and could also be utilized as a low-altitude platform for other sensors. © 2011 Taylor &amp Francis.

Many active volcanoes are distributing in east Asia, however, most of them are not well monitored. Monitoring volcanoes is a key issue for disaster mitigation, as well as scientific studies. It is not realistic to install ground-based instruments to all of the active volcanoes, thus virtually satellite remote sensing is the only way to monitor them scattering in such wide areas. We developed a monitoring system based on Moderate Resolution Imaging Spectroradiometer (MODIS) and Multi-functional Transport Satellite (MTSAT) data, and are currently monitoring 147 active volcanoes in this area. The observation results are uploaded on the website in real or near-real time, after acquisition of observation data. Based on these data, we analyzed details of eruption processes occurred at Asama and Sarychev Peak in 2009, in conjunction with high resolutions images or ground-based observations. We plan to improve the current monitoring system utilizing the advanced observation capabilities of GCOM-C1 SGLI launching in 2014. The SGLI 1.6 μm and 11 μm channels have 250 m resolution, which is four times higher than that of 3.7 μm and 11 μm channels of MODIS. These high resolution channels enable to operate more sensitive thermal monitoring of volcanic activities.

対象物を多方向から放射観測することにより，その反射輝度の変化を表したものを方向別反射特性（BRDF）と呼ぶ。近年衛星リモートセンシングの分野では，BRDFを観測し，それを用いた詳細な植生分類や，樹高などの三次元的な植生パラメータの推定が期待されている。そこで本年度の研究では，八ヶ岳のカラマツ林のタワーにてBRDFの日変化と季節変化の観測を行った。 同一日に複数回BRDFの観測を行った結果，BRDFは観測時の太陽天頂角に依存して日変化した。BRDFの特徴を現す指標HDS（Lacaze et al.,2002,）の値と観測時の太陽天頂角には線形の関係が見られた。また，観測時の太陽天頂角が36°のものを選出し，晩夏～晩秋のHDSの季節変化を検討した結果，秋にかけてHDSの値は減少していき，BRDFの季節依存性が分かった。

A judicious combination of spectral and spatial surface information can improve the understanding of the vegetation optical variability and typological differentiation. The objective of this study is to evaluate the potential of airborne spectral radiation and digital imagery data for vegetation canopy classification and the impact of canopy texture on the vegetation-solar radiation interaction. To conduct the study, two multispectral radiometers with wavelengths ranging from 350 to 1050nm and a fine pixel digital camera are used. One of the radiometers is positioned close to the digital camera, and, both instruments are carried by a radio-controlled helicopter flying above the canopy of a boreal forest of the northern Japanese island of Hokkaido. Analyses of the canopy reflectance signature show a clear species differentiation in the vegetation of the area and give an evaluation of the canopy radiation capacities. The bamboo grass species have the highest reflectance and the needle-leaf species the lowest. To understand the physical factors associated with the reflectance-species typological relationship, textural features are extracted from digital images, by using colour discrimination techniques. The features estimated are the brightness intensity of the canopy, the amounts of gaps and shadows, the degree of heterogeneity of light scattering, and the green vegetation fraction. Then, the relationship between these individual properties and reflectance is examined. The results obtained show that reflectance decreases with increasing amount of gaps and shadows and, increases with the brightness intensity and more importantly with light scattering heterogeneity of the canopy. This heterogeneity effect, derived from the vegetation luminance distribution is examined through three methods. The most elaborate among these methods is the semivariogram analysis. Results of this analysis show that the range of the semivarioragram reflects well enough the average size of the plants (short range for the bamboo grass and large range for the needle-leaf species). The needle-leaf species have the highest variability, i.e. are the most heterogeneous light scatterers, while the bamboo grass species are the least variable. The scale of variability of the distribution of luminance differs according to the species: it is dominated by macrovariability in the needle-leaf, and microvariability in the bamboo grass and the broadleaf. The needle-leaf species' high spatial heterogeneity of light scattering would reduce the measured canopy bi-directional reflected radiation and enhance the transmission of this radiation towards lower vegetation levels through a multiscattering radiation process.

The Fourth Assessment Report of IPCC predicted that global warming is already happening and it should be caused from the increase of greenhouse gases by the extension of human activities. These global changes will give a serious influence for human society. Global environment can be monitored by the earth observation using satellite. For the observation of global climate change and resolving the global warming process, satellite should be useful equipment and its detecting data contribute to social benefits effectively. JAXA (former NASDA) has made a new plan of the Global Change Observation Mission (GCOM) for monitoring of global environmental change. SGLI (Second Generation GLI) onboard GCOM-C (Climate) satellite, which is one of this mission, provides an optical sensor from Near-UV to TIR. Characteristic specifications of SGLI are as follows 1) 250 m resolutions over land and area along the shore, 2) Three directional polarization observation (red and NIR), and 3) 500 m resolutions temperature over land and area along shore. These characteristics are useful in many fields of social benefits. For example, multi-angular observation and 250 m high frequency observation give new knowledge in monitoring of land vegetation. It is expected that land products with land aerosol information by polarization observation are improved remarkably. We are studying these possibilities by ground data and satellite data.

For a comprehensive vegetation monitoring and/or management, a good understanding of the distribution of the solar radiation energy among components of this vegetation is needed. The energy received by the vegetation is measured by spectroradiometers either at satellite elevations or near the ground (in situ measurements). In this study, in situ, radiometric data and laser scanning techniques are combined, in order to evaluate the contribution of the vegetation structure to the variability of canopy reflectance. Advanced processing laser techniques are not only an efficient tool for the generation of physical models but also give information about the vertical structure of canopies (height, shape, density) and their horizontal extension. To conduct this study, airborne multispectral radiation data and, laser pulse returns are recorded from a low flying helicopter above the vegetation of a boreal forest. These measurements are used to derive canopy optical and structural variables. The impact of the canopy 2-dimensional structural variability on the distribution of the solar radiation reflected by plants of this area is discussed. The results obtained show that the laser technology can be used for the selection of the most appropriate configuration of radiation measurements, and optimization of canopy physical characteristics, in future airborne missions.

A study was conducted in Chiba, Japan, to validate Moderate Resolution Imaging Spectroradiometer (MODIS) albedo products by taking the field measurements of shortwave band albedos in paddy fields. A large difference in spatial scale, from field-measured point data to 1-km resolution, complicates the validation process. To assess such effect of different spatial scales, Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and Enhanced Thematic Mapper Plus (ETM+) data were used. Spatial scale effects on the albedo were examined from three viewpoints: 1) comparison between point-based albedo and mean of albedo in homogeneous area; 2) comparison between point-based albedo and 1-km aggregated albedo; and 3) assessment of semivariogram of albedo in homogeneous area. In implementation of viewpoint 2), Liang's regression model was applied to convert ASTER reflectance into shortwave band albedo. The I-km ASTER albedo was estimated using the point spread function, and in the same manner, I-km ETM+ albedo was estimated. All results represent that an area around the measurement site can be assumed to be homogeneous, indicating negligible effects of spatial resolution difference during most of the periods. Comparison of ground-point-based albedos with MODIS actual albedo, estimated from MODIS black-sky albedo, white-sky albedo, and a fraction of diffuse skylight, showed that the accuracy of MODIS albedo products for paddy fields in Japan is within approximately 0.026 by absolute value (root-mean-square error) and 15.1% by relative value.

プロジェクト2衛星データによるユーラシア大陸の植生3次元構造の変遷を中心とする表層・植生・土地被覆変動の研究とデータ解析・処理手法、検証データ観測手法の研究 2-2

Estimating evapotranspiration (ET) is of the highest importance for understanding and eventually intervening in the water cycle of natural systems. ET is one of the major factors influencing climate change at local, regional and global levels. In this study, a surface energy balance method, which combines meteorological observations with spectral data derived from remote sensing measurements, was used to estimate the ET. The Surface Energy Balance Algorithm for Land (SEBAL) has been applied to Landsat+ETM and NOAA-AVHRR sensors for the estimation of ET in Mongolia, a large arid and semi-arid region with homogeneous surface conditions, on 10 August 1999. Actual ET was computed during satellite overpass and integrated for 24-h on pixel-by-pixel basis for daily ET distribution. As a result, a daily ET map over the arid and semi-arid region of Mongolia was analysed some observation data, such as radiations and surface temperature, was compared with estimated data. © 2005 Taylor &amp Francis.

自動分類の問題には, コンピューターが理解できるような植生のカテゴリを記述する手法やそのカテゴリにしたがって分類を行うことなどがある。分類手法には, 教師付き分類法と教師無し分類法の二つがある。これらの手法には自動的に植生のスペクトル特性とカテゴリを結びつける手法がない。著者らはGASC手法をべースとする画像不変式を見出した。<BR>本研究ではその不変式を用いて, ベトナム南部 (Ninh Thuan, Binh Thuan and Lam Dong provinces) においてGLIシミュレーションデータを用いて研究を行った。その結果, 処理時間を短くすることができた。また, この手法で高精度に分類を行うためには広範囲のグランドトゥルースデータが必要であることがあきらかとなった。

1.3データベース研究部門 1.3.9

[ABSTRACT] An automatic selection method of reasonable training areas for supervised texture classification is proposed. This method is based on both the texture representation method using rotation invariant moments and the genetic algorithm (GA). In the proposed method, first, sets of candidates for each texture category are roughly given. The chromosomes for the GA are formed by the indexed candidates and their extent. The fitness function for the GA is obtained by the mixed texture model and quantified pure textures. In this paper, summary of the texture representation method and comparison of texture discrimination ability with the conventional Haralick's method are described. Furthermore, the details of the proposed automatic selection method are presented, and some simulation results are also shown.

極軌道衛星NOAAのAVHRRセンサのデータには, HRPT, LAC, GACの3種類がある。HRPT, LACの地上解像度は1.1kmでGACのそれは約4kmである。GACは低解像度ながら, 他の2つと違って地球全体のデータを地上で受信することができるため, 1982年から現在に至るまでの15年にわたる毎日の全球データが蓄積されている。これまで何度か, このGACを基データとして使用した全球・時系列のクラウドフリーコンポジットデータの作成が米国において行われてきた。しかしながら, それらのコンポジットデータでは, 生データを地図座標系ヘリサンプルする処理においてシステム補正のみを用いているため, 位置精度が非常に悪く, コンポジット処理で地理的に異なった位置のピクセルが時系列化されるという問題点があった。そこで, そのリサンプリング処理でGCPを使った精密幾何補正をおこない, これまでにないクラウドフリーコンポジットデータを自動作成するためのシステム開発を行った。自動処理を行うためにはGCPの自動選択・画像マッチングを用いた対応点検出が必要であり, これらを行うことが非常に難しいために, これまでシステム補正のみが採用されてきた経緯がある。我々はシステム開発においてこの問題を解決し自動処理システムを完成することができた。生成されるコンポジットデータは地上解像度4kmと, GACの解像度と同等であり, この点でもこれまでになかったデータセット作成が可能になった。本システムの特徴は, リサンプリングという時間のかかる処理を, 1パスごとに1台のPCを割り当て, 1日分15パスの処理を並行して行うことである。ギガバイト単位の大量データのファイルI/Oが生じるプロセスでは, 高速な並列計算機を用い