In repeat-pass interferometric synthetic aperture radar (In-SAR), the coherence loss derived from temporal decorrelation can be modeled using an exponential decay function. Many researches have investigated the relationships between decay parameters over various land-cover types for C-band SAR satellites. Although ALOS PALSAR provides many L-band SAR images, its revisit time was insufficient for understanding the rapid exponential drop in temporal decorrelation. Due to the lack of an appropriate archived dataset, analyzing the decay parameters for L-band SAR was a challenging task. In this study, we processed four-year interferometric pairs derived from ALOS-2 PALSAR-2 full-aperture ScanSAR images. This dataset has a 14 day minimum temporal interval. We selected Nairobi, Kenya, in the middle of the East African rift, as the study area and investigated how decay functions differ over multiple land-cover types. In addition, we compared decay models for Sentinel-1 interferograms. This study demonstrates that the decay parameters of L-band coherence completely differs across forest types, whereas C-band coherence is low coherence regardless of forest type.

The translation from Synthetic Aperture Radar (SAR) to optical imagery is vital in remote sensing. Nevertheless, this translation encompasses profound challenges because of the limited information in SAR data compared to optical images. While progress has been achieved through frameworks like pix2pix, the room for enhancement remains plentiful. We have assembled a unique dataset by harmonizing Sentinel-1 and Sentinel-2 satellite images from the SEN12MS dataset, accompanied by JAXA`s High-Resolution Land-Use and Land-Cover (LULC) Map of Japan. Our study proposes an efficient translation framework that modifies the original pix2pix architecture, by modifying the discriminator architecture to a new one that can be trained with LULC, which enables the supervision of the generator using the LULC map. Our method significantly supersedes the conventional pix2pix, indicating LULC discriminator may improve a result with a conventional one.

Deforestation rates in Brazilian Amazon continue to increase and addressing deforestation requires coordinated action at multiple levels, as well as international cooperation despite various strategies and action at multi-levels. To promote such strategies and action, this paper proposed a recurrent deep learning-based model, which predicts monthly deforestation events (will occur or not) at 1 km × 1 km meshes. The model was trained and evaluated by using about seven-year deforestation data (August 2016–September 2023) of eight areas in the Brazilian Amazon from the Real-Time Deforestation Detection System. By using the most recent two-year data, it was clear that the monthly prediction has a reasonable performance for recall, but very low precision. Thus, future work in the monthly prediction of deforestation should focus on refining model accuracy by incorporating more diverse data sources and improving temporal resolution.

Interferometric SAR time series analysis using multilooked interferograms has measured ground deformation with high accuracy over large areas, but is known to be contaminated by a systematic bias when using only short-term interferograms. This bias, also referred to as a "fading signal", has been investigated using the abundant time series data from Sentinel-1 C-band SAR, and soil moisture and biomass changes have been suggested as two possible causes for the observed bias. Several correction methods have been proposed to mitigate the phase bias, assuming a periodic observation strategy. The phase bias is supposed to increase with the wavelength, but the bias in ALOS-2 L-band SAR has not been investigated in detail because the data distribution policy restricts its users. In addition, these correction methods are not suitable for the infrequently observed data such as ALOS-2. In this paper, we investigated the systematic bias in ALOS-2 multilooked interferograms with respect to various interferometric pairs. We also corrected the phase bias using the noise-filtering technique. The bias in ALOS-2 multilooked interferograms showed >3 mm/year with an average temporal baseline of 144 days, which is equivalent to that of Sentinel-1 with the temporal baseline of 78 days. That is, the bias effect toward the temporal baseline was larger in ALOS-2 than in Sentinel-1. The noise-filtering technique could mitigate the bias with less pairs regardless of land-use.

In the aftermath of large-scale natural disasters, the accuracy of building damage detection (BDD) is of critical importance. Post-disaster high-resolution (post-HR) remote sensing imagery is fundamental for BDD; however, prompt acquisition of such imagery remains a significant challenge. To address this issue, we introduce a novel plug-and-play feature fusion (FF) module. This module, strategically situated between a pre-trained super-resolution (SR) model and a BDD model, ingeniously combines features from both pre-disaster high-resolution (pre-HR) and super-resolved post-disaster remote sensing imagery. The proposed approach is designed to maximize the utilization of pre-HR images, thereby enhancing BDD accuracy. Experimental validation confirms that this improvement in accuracy is attributable to the pragmatic extraction of features from pre-HR imagery, not just an increase in model complexity. Consequently, our approach holds substantial promise for real-world post-disaster scenarios and lays a solid foundation for future BDD research, demonstrating potential improvements in both efficacy and practicality.

In this study, we proposed a new method for flat-flied calibration of a sensor in space, which is basically difficult to conduct in space. In this method, an observation of the Moon with cross-track scan and its simulation are keys to achieve conducting the calibration. By applying the proposed method to GOSAT-2/CAI-2’s Moon observations with cross-track scan, we successfully obtained sensitivity difference among pixels in a sensor quantitatively, and it is consistent with that in a ground test before launch. In addition, by comparing results from different observations, it is possible to assess temporal variations of sensitivities in various positions in a sensor, which is a new information.

Building damage detection (BDD) with satellite images has been frequently adopted as an essential reference for post-disaster rescue, whereas its timeliness is significantly impacted by the long revisit time of high-resolution remote sensing satellites. Therefore, a reliable super-resolution method which is optimized for accurate and detail BDD is fundamental one for advancing the BDD analysis even when we can use only low-resolution images after a disaster. Based on Super-Resolution Generative Adversarial Network (SRGAN) and U-Net convolutional network, an efficient and novel BDD framework is proposed in this paper for obtaining upsampled BDD results from low-resolution post-disaster images. We trained the framework using two disasters from the xBD dataset and tested three different structures. The results show that our training structure based on an end-to-end framework successfully generated super-resolution BDD maps from low-resolution images, which performed significantly better than those from a two-stage training structure.

Radiometric calibration with the Moon (called the lunar calibration) is a promising method for evaluating the performance of instruments onboard satellites. In particular, the lunar calibration can provide inflight radiometric calibration opportunities for nano/microsatellites after the launch without any special equipment. In this study, we have conducted more than one year of Moon observations with a microsatellite named RISESAT. The performance of an onboard multispectral sensor OOC is investigated by using the ROLO and SELENE/SP Moon models. Our lunar calibration revealed the dependence of the sensor sensitivity on the instrument temperature, and no sensitivity degradation (< 1 %) can be confirmed for all four channels after the correction of the temperature dependence. Lunar calibration also revealed the bluing trend in the OOC’ s inter-band ratio, which indicates a 15% alternation in the sensitivity at maximum. This bluing trend was validated by the Railroad Valley Playa observations as comparing the Landsat-8/OLI data.

Since the global spread of COVID-19 in 2020, in order to reduce infections, the movement of people has been severely restricted. As a result, the economic environment in commercial aviation suffered an unprecedented impact. Therefore, it becomes important to study the impact of COVID-19 on commercial aviation. In this study, in order to understand the changing trend of the number of airplanes as an index of the airport activity, we applied a method that utilizes convolutional neural networks to effectively detect airplane in Sentinel-2 images. From the detection, we successfully obtained the changing trend of the number of airplanes in important airports around the world since the outbreak of COVID-19, and found different changing trends in different areas that may reflect different reactions to COIVD-19 situation in each country.

Radiometric calibration with the Moon (called the lunar calibration) is a promising method for evaluating the performance of instruments onboard satellites orbiting around the Earth. In particular, the lunar calibration can provide inflight radiometric calibration opportunities for nano/microsatellites whose priorities are technical demonstrations and scientific observations during rather a short lifetime. In this study, we have applied the lunar calibration to a recently launched microsatellite named RISESAT. RISESAT had observed the Moon every month for a half year after the launch with an onboard instrument OOC. By simulating the Moon brightness for each observation based on the ROLO and SELENE/SP Moon models, we succeeded to measure <; 2 % of a small degradation in four months. Further comparison of the observation and the simulation irradiance of the Moon revealed the bluing trend in the OOC's inter-band ratio.

Since the huge earth quake in 2011 in Japan, the Japanese government has strongly encouraged growth of renewable energy use. As a result, even we focus on only mega-solar photovoltaic (PV) power plants (> 1,000 kW), the amount of the power generation significantly increased. Because such rapid increasing of the PV power plants changes land use widely, verification of the spreading of the PV power plants should be essential for assessing economic/environmental issues. In this study, to verify the increasing of the PV power plants based on observations, we applied a method that can detect the PV power plants efficiently with a convolutional neural network to Landsat-8 images. From the detection, we successfully identified the increasing of the area of the PV power plants quantitatively at the same time identifying the location of each PV power plant in many prefectures in the capital region of Japan.

Various saliency detection algorithms from color images have been proposed to mimic eye fixation or attentive object detection response of human observers for the same scenes. However, developments on hyperspectral imaging systems enable us to obtain redundant spectral information of the observed scenes from the reflected light source from objects. A few studies using low-level features on hyper-spectral images demonstrated that salient object detection can be achieved. In this work, we proposed a salient object detection model on hyperspectral images by applying manifold ranking (MR) on self-supervised Convolutional Neural Network (CNN) features (high-level features) from unsupervised image segmentation task. Self-supervision of CNN continues until clustering loss or saliency maps converges to a defined error between each iteration. Finally, saliency estimations is done as the saliency map at last iteration when the self-supervision procedure terminates with convergence. Experimental evaluations demonstrated that proposed saliency detection algorithm on hyperspectral images is outperforming state-of-the-arts hyperspectral saliency models including the original MR based saliency model.

ASTER, which is a multi-band pushbroom sensor onboard Terra, has provided worthful multiband images for approximately 20 years successfully since Terra's launch in 1999. For such a long operation, maintaining accurate radiometric calibration is an important challenge to guarantee the reliability of the satellite products. In April 2003 and August 2017, ASTER observed the Moon (and deepspace) for conducting radiometric calibration with the Moon, which is called as lunar calibration. Through the lunar calibration, temporal variation of relative sensor sensitivity can be measured accurately (the order of 0.1 %), which is reasonably accurate to improve calibration results from other methods. The lunar calibration was performed to ASTER's visible and near infrared bands (Band 1, Band 2, and Band 3N) by comparing the observed brightness with simulated lunar brightness. 3, 5, and 6% sensitivity degradation were observed in Band 1, Band 2, and Band 3N of ASTER, respectively, which are basically consistent with vicarious calibration, though still further discussion is needed for combining calibration results from different methods.

UNIFORM-1 is a small satellite that helps in the prevention of severe fire hazards by detecting the wildfire at an early stage. UNIFORM-l is equipped with an un-cooled bolometer imager (BOL), which observes temperature distributions of the earth surface. Uniquely, BOL is not equipped with any shutter mechanism to reduce the risk associated with the shutter that impacts the ability to conduct any observation in space. As a trade-off of shutter exclusion, BOL saw the Sun in its field of view (FOV) many times and the sunlight damage caused large brightness temperature error (60 K) in BOL images. To reduce the error, we have proposed a new calibration procedure utilizing deepspace observations, which enables the measurement of the increase in digital count offset due to the sunlight in a BOL image. After the correction, we successfully reduced the damage, and the relative accuracy of brightness temperature improved to more than 2 K, which is enough to detect a wildfire with a size smaller than one-pixel scale.

This paper presents an efficient object detection method from satellite imagery. Among a number of machine learning algorithms, we proposed a combination of two convolutional neural networks (CNN) aimed at high precision and high recall, respectively. We validated our models using golf courses as target objects. The proposed deep learning method demonstrated higher accuracy than previous object identification methods.

Many works have been done on salient object detection using supervised or unsupervised approaches on colour images. Recently, a few studies demonstrated that efficient salient object detection can also be implemented by using spectral features in visible spectrum of hyperspectral images from natural scenes. However, these models on hyperspectral salient object detection were tested with a very few number of data selected from various online public dataset, which are not specifically created for object detection purposes. Therefore, here, we aim to contribute to the field by releasing a hyperspectral salient object detection dataset with a collection of 60 hyperspectral images with their respective ground-truth binary images and representative rendered colour images (sRGB). We took several aspects in consideration during the data collection such as variation in object size, number of objects, foreground-background contrast, object position on the image, and etc. Then, we prepared ground truth binary images for each hyperspectral data, where salient objects are labelled on the images. Finally, we did performance evaluation using Area Under Curve (AUC) metric on some existing hyperspectral saliency detection models in literature.

We demonstrate how to utilize a two-dimensional image sensor onboard a satellite for determining its attitude. The method is based on image matching between satellite images and Earth surfaces. To achieve this, image feature extraction and robust estimation techniques are employed. Experimental results showed that the accuracy of attitude determination is about 0.02° if the satellite position has been precisely determined. The contribution of this paper is not on technical novelty, as the method has already been described in [1], but on additional examples illustrating the utility of the method.

In recent years, more and more small satellites have been launched and operated for various purposes. Because of their sever restrictions of payload weight and cost, a radiometric calibration approach which does not need any special instrument has been desired. In this study, we demonstrate a radiometric calibration approach for Hodoyoshi-1, a Japanese small satellite, based on Moon observations utilizing SLENE/SP Lunar surface reflectance model. The model can simulate any satellite Moon observation considering observation geometry, and can be used for evaluating the consistency of the observed Moon brightness, which allows us to measure a time variation of a sensor sensitivity. Through Hodoyoshi-1's three Moon observations, gradual increasing in relative sensor sensitivities were confirmed.

The thermal infrared camera with the uncooled microbolometer array based on commercial products has been developed in a laboratory of a Japanese university and mounted to a 50-kg class small satellite specialized for discovering wildfire. It has been launched in 2014 and successfully detected considerable hotspots not only wildfire but also volcanoes. Brightness temperature derived from observation has been verified, and the scale of observed wildfire has been provisionally presumed; the smallest wildfire ever detected has a flame zone less than ~300 m2and the fire radiative power = ~35.4 mW. It is 1/30th the size of the initial requirement estimated in the design process. Our thermal infrared camera developed in a short time with low cost has attained enough ability to discover small wildfire which is suppressive at initial attack.

The hyperspectral imager of the HISUI (Hyperspectral Imager Suite) project will be deployed and operated in the International Space Station (ISS) Japanese Experiment Module-Exposed Facility (JEM-EF) in FY2018. The pre-launch calibration is conducted using a large integrating sphere traceable to the SI, and after the launch, the onboard calibration will be carried out using onboard calibration unit including a lamp. The vicarious calibration will be conducted over the dry lake sites located both in northern hemisphere and in southern hemisphere, and we may also use automated calibration facilities due to the effects of orbital characteristics for the ISS. The HISUI hyperspectral imager will be cross-calibrated using forthcoming hyperspectral sensors flying on ISS. The data associated with these calibrations will be archived in the Calibration Data Archive System (CDAS), and the HISUI calibration working group plans to assess HISUI radiometric performance and produces radiometric (geometric) calibration databases on CDAS, which are required in Level 1 processing.

The Hyperspectral Imager Suite (HISUI) is a future hyper-spectral and multispectral sensor developed by Japanese Ministry of Economy, Trade, and Industry (METI). Radiometric calibration of satellite sensor system is very important for the higher level products with high quality. This paper mainly shows the vicarious calibration and cross-calibration for assessment of HISUI radiometric performance in HISUI Calibration Working Group.

The Hyperspectral Imager Suite (HISUI) is the Japanese next-generation Earth observation project, and is being developed by Japanese Ministry of Economy, Trade, and Industry (METI). We conduct the vicarious calibration for HISUI instrument as the pre-launch activities in summer in the southern hemisphere. Recent our activities focus on vicarious calibration experiment over Lake Lefroy, which is a large salt lake in southern Western Australia. Cross-calibration for HISUI is also discussed. This paper shows both of vicarious calibration and cross-calibration for HISUI instrument.

We have developed a method for radiometric calibration of HISUI's hyper and multi-spectral sensors using a lunar reflectance model developed from SELENE SP data, which involves lunar surface reflectance and photometric properties. For evaluating the utilization of the model, we simulated a lunar observation conducted by ASTER of its three visible and infrared bands and confirmed the model describes lunar surface photometric properties correctly because correlation coefficients of observed and modeled radiance exceed 0.99 for all bands. Although absolute radiance shows some discrepancy between the observed and the simulated Moon in visible band, the model is, at least, useful to evaluate relative degradation of sensors.