Distribution-Aware Interactive Attention Network and Large-Scale Cloud Recognition Benchmark on FY-4A Satellite Image

A method in near-surface estimation of air temperature (NEAT) in times following the satellite passing time using MODIS images

Acid mine drainage (AMD) is a severe environmental issue associated with the generation of finely milled rock waste containing high concentrations of sulfide minerals and potentially toxic elements (PTEs) during mining activities. The formation of secondary minerals, such as sulfates and iron oxyhydroxides, results from sulfide oxidation and subsequent acid neutralization by carbonate and silicate minerals, making them key indicators of AMD. Efficient identification of these minerals is crucial for monitoring their impact on soils.This study compares the capabilities of Landsat-09, ASTER, and Sentinel-2 satellite images in identifying Jarosite, Goethite, Ferrihydrite, Anhydrite, and Gypsum (associated with AMD) using the "Spectral Angle Mapping" (SAM) technique. SAM is a spectral analysis method that classifies materials based on the angle between spectral vectors corresponding to their spectral signatures.The evaluated satellite images were selected based on their spatial, spectral, and temporal resolution. Their strengths and limitations in detecting the selected secondary minerals were assessed using the SAM technique in ENVI software. The algorithm was trained with spectral signatures ranging from 0.4 to 2.5 micrometers, obtained from the USGS and ASTER spectral libraries. Landsat-09 offers moderate resolution and global coverage; ASTER excels in shortwave infrared capabilities but lacks recent satellite imagery for current analyses; and Sentinel-2 combines high resolution with a broad spectral range and biweekly temporal resolution, with continuous image acquisition to date.The results show significant differences in each sensor's ability to identify the minerals of interest. Sentinel-2 demonstrated high accuracy due to its spatial resolution and specific spectral bands. Conversely, ASTER was unable to precisely delineate pixels associated with the requested minerals. Lastly, Landsat-09 showed limitations in mineral identification using this technique due to the sensor’s spatial resolution. This study highlights that spatial resolution is the most critical factor in selecting satellite imagery for SAM applications. Thus Sentinel-2, with the highest spatial resolution (10 m) achieved superior results in identifying AMD-related minerals.This study provides guidance for selecting satellite sensors based on spatial and spectral resolutions in studies aimed at mineral identification using SAM. It contributes to the development of more efficient strategies for environmental management, mineral exploration, and energy resource studies, among other applications.

The fusion of unmanned aerial system (UAS) and satellite imagery has emerged as a pivotal strategy in advancing precision agriculture. This review explores the significance of integrating high-resolution UAS and satellite imagery via pixel-based, feature-based, and decision-based fusion methods. The study investigates optimization techniques, spectral synergy, temporal strategies, and challenges in data fusion, presenting transformative insights such as enhanced biomass estimation through UAS-satellite synergy, improved nitrogen stress detection in maize, and refined crop type mapping using multi-temporal fusion. The combined spectral information from UAS and satellite sources confirms instrumental in crop monitoring and biomass estimation. Temporal optimization strategies consider factors such as crop phenology, spatial resolution, and budget constraints, offering effective and continuous monitoring solutions. The review systematically addresses challenges in spatial and temporal resolutions, radiometric calibration, data synchronization, and processing techniques, providing practical solutions. Integrated UAS and satellite data impact precision agriculture, contributing to improved resolution, monitoring capabilities, resource allocation, and crop performance evaluation. A comparative analysis underscores the superiority of combined data, particularly for specific crops and scenarios. Researchers exhibit a preference for pixel-based fusion methods, aligning fusion goals with specific needs. The findings contribute to the evolving landscape of precision agriculture, suggesting avenues for future research and reinforcing the field’s dynamism and relevance. Future works should delve into advanced fusion methodologies, incorporating machine learning algorithms, and conduct cross-crop application studies to broaden applicability and tailor insights for specific crops.

Due to the limitations of hardware technology, satellite sensors cannot obtain images with high temporal, spatial, and spectral resolutions at the same time. Current spatiotemporal fusion methods try to solve the contradiction between temporal resolution and spatial resolution, which cannot achieve good reconstruction accuracy partly because the data sources are from heterogeneous platforms with long chains difficult to be modeled. Different from the crossing-platform fusion, this work proposes to improve the spatial and temporal resolutions on a single platform. For the 2-m panchromatic images, 8-m multispectral images, and 16-m wide-field-view images captured by the Gaofen-1 satellite, our goal is to produce 2-m multispectral images with high temporal resolutions. Two convolutional neural networks are built to solve this spatiotemporal-spectral fusion issue with pansharpening and spatiotemporal fusion in serial. In the validation stage, the 2-m multispectral images are built and evaluated with the panchromatic images and 8-m multispectral images. The digital and visual evaluations show that our method can produce visually acceptable fusion quality, which may enhance the feasibility of the Gaofen-1 data.

Introduction: Neural crest (NC) cells are involved in cardiac development, and NC defects account for over 10% of all congenital heart disease. The migratory and morphogenic patterns of these NC cells are not well understood because conventional optical microscopes struggle to acquire images of live embryos with sufficient temporal and spatial resolution without causing major photo-damage. To address these issues, the lab has designed and built a light-sheet fluorescence microscope (LSFM) that is capable of acquiring dual-channel signals with high spatial and temporal resolution. Hypothesis: We hypothesize that the LSFM system will capture heart tube development at a cellular level and show the time point and location at which the NC cells integrate with the heart tube. Methods and Results: The transgenic line that is manipulated for this study is the Tg(neural crest:nfsB-mCherry, cardiac: eGFP) in which the heart tube fluoresces green while NC-derived cells fluoresce red. 10 zebrafish were imaged using the 3-D time-lapse LSFM setup from 15 hours post fertilization (hpf) to 36 hpf. The imaging data was qualitative and consistent across fish. Fig. 1 (a) shows mCherry NC-derived cells (red) integrating with eGFP cardiomyocytes (green) and areas of colocalization (yellow). Fig 1 (b) and (c) show the time course of the NC cell migration into the heart tube. By 26-27 hpf, the NC cells contacted the dorsal surface of the heart tube. By the 29 hpf, the NC-derived cells integrated into the heart tube and expressed nuclear eGFP, which is visualized by the yellow fluorescing portions of the image. Conclusions: We conclude that the LSFM was successful in acquiring distinct signals from the heart tube cells and the NC-derived cells. The temporal and spatial resolution was sufficient to observe NC migration at the single cell level and map out exactly where along the heart tubule the NC-cells entered.

Parameterization of air temperature in high temporal and spatial resolution from a combination of the SEVIRI and MODIS instruments

Progress in modern neuroscience critically depends on our ability to observe the activity of large neuronal populations with cellular spatial and high temporal resolution. However, two bottlenecks constrain efforts towards fast imaging of large populations. First, the resulting large video data is challenging to analyze. Second, there is an explicit tradeoff between imaging speed, signal-to-noise, and field of view: with current recording technology we cannot image very large neuronal populations with simultaneously high spatial and temporal resolution. Here we describe multi-scale approaches for alleviating both of these bottlenecks. First, we show that spatial and temporal decimation techniques based on simple local averaging provide order-of-magnitude speedups in spatiotemporally demixing calcium video data into estimates of single-cell neural activity. Second, once the shapes of individual neurons have been identified at fine scale (e.g., after an initial phase of conventional imaging with standard temporal and spatial resolution), we find that the spatial/temporal resolution tradeoff shifts dramatically: after demixing we can accurately recover denoised fluorescence traces and deconvolved neural activity of each individual neuron from coarse scale data that has been spatially decimated by an order of magnitude. This offers a cheap method for compressing this large video data, and also implies that it is possible to either speed up imaging significantly, or to “zoom out” by a corresponding factor to image order-of-magnitude larger neuronal populations with minimal loss in accuracy or temporal resolution.

For more than four decades, geostationary weather satellites have been used successfully by NOAA's National Weather Service for daily weather forecasting and for tracking storm events. At the same time, polar orbit systems, such as NOAA/AVHRR, SeaWIFS and Landsat TM, have provided valuable information about slowly varying features and processes in the open ocean and on land. However, coastal and estuarine applications of remote sensing require significantly finer temporal, spatial, and spectral resolution than open ocean studies. Current polar orbit satellites lack the temporal resolution, while present geostationary orbit satellites have limited spatial and spectral resolution. Geostationary satellites can provide images of clouds, ocean and land surfaces very frequently, since they are positioned above a selected point on the equator and continuously view the same one third of the Earth. However, in order to have the same angular velocity as the Earth, they must be located in an orbit 23,000 miles above the equator, which limits their spatial resolution. NOAA, with support from NASA and the Navy, is developing GOES-R, a geostationary satellite. GOES-R will include a Hyperspectral Environmental Suite for Coastal Waters (HES-CW) and atmospheric sounding sensors which should provide the resolution levels required for coastal and estuarine observations. To achieve this, full use will be made of the fact that a geostationary platform offers a very high sampling frequency and long integration times, allowing for high signal-to-noise (S/N) ratios. Thus, useful trade-offs can be made between S/N, number of spectral bands, and sampling frequency, even though the spatial resolution is constrained to 150-300 m at the equator by system design considerations, such as mirror size, stability and pointing of the spacecraft. The objective of this paper is to illustrate coastal and estuarine resolution requirements for typical coastal features and processes such as: 1) chlorophyll concentration and coastal productivity, 2) harmful algal blooms and eutrophication, 3) current measurement by feature tracking, 4) oil slick drift and capture by estuarine fronts, 5) dynamic behavior of turbidity plumes and ocean-dumped waste, 6) propagation of internal waves on the continental shelf, 7) ice cover for coastal shipping, 8) coastal flooding and other catastrophic events, and 9) coastal ecosystem and wetland habitat changes. These resolution requirements are then compared to current GOES-R specifications.

Coastal change assessments have significant socioeconomic, environmental, and infrastructure implications due to the extensive impacts of climate change, such as rising sea levels, the increasing frequency and intensity of storms, as well as the consequences of human intervention.  Satellite products have been used to monitor the coast at relatively high resolution (30 m) since the 1970s through the Landsat program. However, the arrival of the EU Copernicus/Sentinels in 2014 introduced a marked increase in coastal applications thanks to the improved spatial and temporal resolutions. The research presented in this study explores the derivation of waterlines from Sentinel 2 imagery and the creation of a novel holistic approach to a validation framework. Specifically, this study aims to: a) explore the inherent waterline errors against reference datasets and begin to establish the overall uncertainty in deriving waterlines from optical satellite imagery; and b) assess the potential of these results and their suitability for coastal change applications. The results indicate an average positional error of approximately 4 meters for Sentinel images in coastal regions by evaluating the Sentinel-2 satellite images with distinct features visible in aerial orthophotography. Subsequently, the horizontal and vertical inaccuracies of the satellite-derived waterlines (SDWL) were further determined by using a GNSS line as a reference dataset. The horizontal assessment was conducted by calculating the average distance between the SDWLs and the GNSS reference lines across eighteen Sentinel-2 images corresponding to the years 2021, 2022, 2023 and 2024. These were analysed, showing a median displacement of 15 meters, and indicating an offshore trend for the satellite-derived waterlines. The vertical assessment, or height error, was computed by comparing the average height of SDWLs (as determined by the average tide gauge heights) with the reference dataset height (as measured by GNSS), resulting in a mean absolute error of 6 cm. The vertical results indicate that the SDWLs’ heights, as measured by the local tide gauges, align well with in situ local height measurements. The results of this study will aid in identifying temporal and spatial scales and resolutions at which Earth Observation products are suitable for coastal management. The initial stages of a validation framework are presented to assess the quality and applicability of satellite-derived waterlines for coastal change monitoring based on specific user requirements. Identifying the sources of error and improving uncertainty models for satellite-derived products enables better decision-making in coastal management. These analyses will demonstrate whether the outcomes remain consistent among satellite images or change according to local environmental conditions. Increasing end-user confidence in the rates of change obtained from available satellite products can provide crucial information in study areas, and at space-time resolutions previously unattainable.

Weather radar networks are an excellent tool for quantitative precipitation estimation (QPE), due to their high resolution in space and time, particularly in remote mountain areas such as the Tropical Andes. Nevertheless, reduction of the temporal and spatial resolution might severely reduce the quality of QPE. Thus, the main objective of this study was to analyze the impact of spatial and temporal resolutions of radar data on the cumulative QPE. For this, data from the world’s highest X-band weather radar (4450 m a.s.l.), located in the Andes of Ecuador (Paute River basin), and from a rain gauge network were used. Different time resolutions (1, 5, 10, 15, 20, 30, and 60 min) and spatial resolutions (0.5, 0.25, and 0.1 km) were evaluated. An optical flow method was validated for 11 rainfall events (with different features) and applied to enhance the temporal resolution of radar data to 1-min intervals. The results show that 1-min temporal resolution images are able to capture rain event features in detail. The radar–rain gauge correlation decreases considerably when the time resolution increases (r from 0.69 to 0.31, time resolution from 1 to 60 min). No significant difference was found in the rain total volume (3%) calculated with the three spatial resolution data. A spatial resolution of 0.5 km on radar imagery is suitable to quantify rainfall in the Andes Mountains. This study improves knowledge on rainfall spatial distribution in the Ecuadorian Andes, and it will be the basis for future hydrometeorological studies.

Timely and accurate monitoring of tree crop extent and productivities are necessary for informing policy-making and investments. However, except for a very few tree species (e.g., oil palms) with obvious canopy and extensive planting, most small-crown tree crops are understudied in the remote sensing domain. To conduct large-scale small-crown tree mapping, several key questions remain to be answered, such as the choice of satellite imagery with different spatial and temporal resolution and model generalizability. In this study, we use olive trees in Morocco as an example to explore the two abovementioned questions in mapping small-crown orchard trees using 0.5 m DigitalGlobe (DG) and 3 m Planet imagery and deep learning (DL) techniques. Results show that compared to DG imagery whose mean overall accuracy (OA) can reach 0.94 and 0.92 in two climatic regions, Planet imagery has limited capacity to detect olive orchards even with multi-temporal information. The temporal information of Planet only helps when enough spatial features can be captured, e.g., when olives are with large crown sizes (e.g., >3 m) and small tree spacings (e.g., <3 m). Regarding model generalizability, experiments with DG imagery show a decrease in F1 score up to 5% and OA to 4% when transferring models to new regions with distribution shift in the feature space. Findings from this study can serve as a practical reference for many other similar mapping tasks (e.g., nuts and citrus) around the world.

Introduction Dynamic MRI may be used to image motion that is needed for accurate radiotherapy treatment planning. The tradeoff between temporal and spatial resolution has been a major issue affecting the use of dynamic MRI. Here we propose to improve temporal resolution by reconstructing high spatial resolution components with deformably registered motion data. Dynamic MRI with increased temporal resolution without scarifying spatial resolution for several sites are presenting. Methods and Materials High and low spatial resolution components of the image typically move similarly, i.e. share most of the motion vector information. The motion vectors can be obtained by deformable registration of the central k‐space, and applied to the high spatial resolution data. This allows one to acquire the peripheral k‐space at a lower temporal resolution than the central k‐space therefore increase the overall sampling rate. We tested the possible gain by this method for 2‐dimensional dynamic MRIs (SIEMENS 3T Verio) of swallowing, breathing and peristalsis. Results The mean error in the displacement vectors obtained using only the central 20% of k‐space vs. using the full k‐space was less than 0.5mm, which reduced to 0.2mm if 50% of the radius is used. Peristalsis motion with sudden changes shared least information between central and peripheral k‐space. The reconstructed images with as low as central 20% of the k‐space radius are very similar to images reconstructed with full k‐space. The resulting probability distance histograms and regions of interest show minimal variation. Conclusion Temporal resolution can be improved considerably without compromising the spatial resolution if the low‐resolution vector information is used to construct high spatial resolution data as demonstrated in this work. The deformable registration accuracy is highly critical for the overall performance of the method. Conflict of Interest: This work is supported partially by Siemens Healthcare.

Remote sensing data represent one of the most important sources for automized yield prediction. High temporal and spatial resolution, historical record availability, reliability, and low cost are key factors in predicting yields around the world. Yield prediction as a machine learning task is challenging, as reliable ground truth data are difficult to obtain, especially since new data points can only be acquired once a year during harvest. Factors that influence annual yields are plentiful, and data acquisition can be expensive, as crop-related data often need to be captured by experts or specialized sensors. A solution to both problems can be provided by deep transfer learning based on remote sensing data. Satellite images are free of charge, and transfer learning allows recognition of yield-related patterns within countries where data are plentiful and transfers the knowledge to other domains, thus limiting the number of ground truth observations needed. Within this study, we examine the use of transfer learning for yield prediction, where the data preprocessing towards histograms is unique. We present a deep transfer learning framework for yield prediction and demonstrate its successful application to transfer knowledge gained from US soybean yield prediction to soybean yield prediction within Argentina. We perform a temporal alignment of the two domains and improve transfer learning by applying several transfer learning techniques, such as -SP, BSS, and layer freezing, to overcome catastrophic forgetting and negative transfer problems. Lastly, we exploit spatio-temporal patterns within the data by applying a Gaussian process. We are able to improve the performance of soybean yield prediction in Argentina by a total of 19% in terms of RMSE and 39% in terms of compared to predictions without transfer learning and Gaussian processes. This proof of concept for advanced transfer learning techniques for yield prediction and remote sensing data in the form of histograms can enable successful yield prediction, especially in emerging and developing countries, where reliable data are usually limited.

Structured illumination microscopy (SIM) is a popular and powerful super-resolution (SR) technique in biomedical research. However, the conventional reconstruction algorithm for SIM heavily relies on the accurate prior knowledge of illumination patterns and signal-to-noise ratio (SNR) of raw images. To obtain high-quality SR images, several raw images need to be captured under high fluorescence level, which further restricts SIM’s temporal resolution and its applications. Deep learning (DL) is a data-driven technology that has been used to expand the limits of optical microscopy. In this study, we propose a deep neural network based on multi-level wavelet and attention mechanism (MWAM) for SIM. Our results show that the MWAM network can extract high-frequency information contained in SIM raw images and accurately integrate it into the output image, resulting in superior SR images compared to those generated using wide-field images as input data. We also demonstrate that the number of SIM raw images can be reduced to three, with one image in each illumination orientation, to achieve the optimal tradeoff between temporal and spatial resolution. Furthermore, our MWAM network exhibits superior reconstruction ability on low-SNR images compared to conventional SIM algorithms. We have also analyzed the adaptability of this network on other biological samples and successfully applied the pretrained model to other SIM systems.

Global surveillance based on remote sensing always has a trade-off between spectral, temporal, and spatial resolutions of the sensor data. Since unique land cover information could be estimated at a good spatial scale than the temporal resolution, the Sentinel sensor, for example, has a revisit of 5 days and gathers images at a 10-20m spatial resolution. The Landsat sensor can only be revisited every 16 days with a spatial scale of 30 m. Fine temporal and spatial resolution data are essential for seasonable observation of dynamic agricultural, environmental, or ecological phenomena. This trade-off problem can be solved by using fusion methods based on the linear differences between images with fine (Sentinel) and coarse (Landsat) spatial resolution. As a result, an image with high spatial resolution can be predicted than acquired from a single Landsat-like sensor. The proposed work will combine Landsat and Sentinel data to bridge the gap between the Landsat image (spatial resolution,30m) and the Sentinel image (temporal resolution,10m or 20m). As a result, detailed investigations of the Earth’s surface and dynamics will be possible. Evaluation metrics – RMSE, SRE, and SSIM are used to compare the predicted image with the Sentinel image of the predicted date. The metrics analysis found that the predicted image is like the reference image with optimal values for all three-evaluation metrics.