Real Satellite Data Research Articles

Democratizing Direct-to-Cell Low Earth Orbit Satellite Networks

The emergent Direct-to-cell Low Earth Orbit (LEO) satellites promise to eliminate cellular coverage dead zones by enabling direct network access to our regular phones and IoTs everywhere on Earth via 4G, 5G, and beyond. However, their large-scale deployment can be impeded by scarce orbital slots in space, a lack of licensed cellular spectrums, and prohibitive capital costs for satellite deployment and operation. This article explores how to enable multi-tenant direct-to-cell LEO satellites as a cost-effective win-win solution for satellite operators, mobile operators, and users. We demonstrate that the state-of-the-art cellular networks' stateful hop-by-hop session-based architecture restricts the LEO satellite multi-tenancy. To remove this technical barrier, we propose MOSAIC [1], an alternative scheme to shift from the hop-by-hop session-based cellular service to pay-as-you-go satellite self-service. MOSAIC lets mobile operators supply their users with one-time digitally signed tokens as "coins" to pay for local satellite access on demand. Any satellite operator's authentic satellite can accept these tokens to serve this user without pre-establishing sessions or contacting mobile operators. Akin to cloud computing, this paradigm lets satellite operators flexibly lease their satellites to more mobile operators or higher revenues and return on investment, and allows mobile operators and users to enjoy competitive satellites for complementary coverage and cost optimizations. Our evaluations with the real satellite data and commodity 3GPP NTN protocol stack validate MOSAIC's scalability and performance advantages compared to the state-of-the-art.

Ocean Satellite Data Fusion for High-Resolution Surface Current Maps

Real-time reconstruction of ocean surface currents is a challenge due to the complex, non-linear dynamics of the ocean, the small number of in situ measurements, and the spatio-temporal heterogeneity of satellite altimetry observations. To address this challenge, we introduce HIRES-CURRENTS-Net, an operational real-time convolutional neural network (CNN) model for daily ocean current reconstruction. This study focuses on the Mediterranean Sea, a region where operational models have great difficulty predicting surface currents. Notably, our model showcases higher accuracy compared to commonly used alternative methods. HIRES-CURRENTS-Net integrates high-resolution measurements from the infrared or visible spectrum—high resolution Sea Surface Temperature (SST) or chlorophyll (CHL) images—in addition to the low-resolution Sea Surface Height (SSH) maps derived from satellite altimeters. In the first stage, we apply a transfer learning method which uses a high-resolution numerical model to pre-train our CNN model on simulated SSH and SST data with synthetic clouds. The observation of System Simulation Experiments (OSSEs) offers us a sufficient training dataset with reference surface currents at very high resolution, and a model trained on this data can then be applied to real data. In the second stage, to enhance the real-time operational performance of our model over previous methods, we fine-tune the CNN model on real satellite data using a novel pseudo-labeling strategy. We validate HIRES-CURRENTS-Net on real data from drifters and demonstrate that our data-driven approach proves effective for real-time sea surface current reconstruction with potential operational applications such as ship routing.

SAR sensing of the atmosphere: stack-based processing for tropospheric and ionospheric phase retrieval

ABSTRACT This paper is intended to summarize the research conducted during the first 2 years of the Dragon 5 project 59,332 (geophysical and atmospheric retrieval from Synthetic Aperture Radar (SAR) data stacks over natural scenarios). Monitoring atmospheric phenomena, encompassing both tropospheric and ionospheric conditions, holds pivotal significance for various scientific and practical applications. In this paper, we present an exploration of advanced techniques for estimating tropospheric and ionospheric phase screens using stacks of Synthetic Aperture Radar (SAR) images. Our study delves into the current state-of-the-art in atmospheric monitoring with a focus on spaceborne SAR systems, shedding light on their evolving capabilities. For tropospheric phase screen estimation, we propose a novel approach that jointly estimates the tropospheric component from all the images. We discuss the methodology in detail, highlighting its ability to recover accurate tropospheric maps. Through a series of quantitative case studies using real Sentinel-1 satellite data, we demonstrate the effectiveness of our technique in capturing tropospheric variability over different geographical regions. Concurrently, we delve into the estimation of ionospheric phase screens utilizing SAR image stacks. The intricacies of ionospheric disturbances pose unique challenges, necessitating specialized techniques. We dissect our approach, showcasing its capacity to mitigate ionospheric noise and recover precise phase information. Real data from the Sentinel-1 satellite are employed to showcase the efficacy of our method, unraveling ionospheric perturbations with improved accuracy. The integration of our techniques, though presented separately for clarity, collectively contributes to a comprehensive framework for atmospheric monitoring. Our findings emphasize the potential of SAR-based approaches in advancing our knowledge of atmospheric processes, thus fostering advancements in weather prediction, geophysics, and environmental management.

Метод оперативной корректуры спутниковой гидрометеорологической информации в районе плавания судна с помощью подспутниковых измерений

The problem of assessing the state of the sea surface by combining satellite and ship measurements is considered. Information about the strength and direction of the wind at the sea surface is important to ensure safe navigation and planning of vessel transit routes. It is of key importance for weather forecasting and excitement. Wind data is traditionally obtained from passing ships and buoys. Ship data covers only limited areas of the World Ocean and is received unevenly, buoy data is highly sparse, which makes it impossible to obtain an adequate picture of the distribution of atmospheric currents. The use of remote sensing satellites has made it possible to increase the density of measurements of wind speed and direction by orders of magnitude, and the relevance of the data obtained. However, satellite measurements can be distorted. In particular, the influence of atmospheric precipitation (rain) can give a two-fold error in measuring wind speed. Strong and weak winds also create problems: in strong winds, the measured data underestimate the true wind speed, in weak winds they overestimate. A method for correcting satellite data on wind speed and direction is proposed. The essence of the method is to use subsatellite ship measurements in conjunction with satellite information. Knowing the weather data measured locally (from the ship), it is possible to calculate regression parameters for correcting satellite data. By applying regressions with calculated parameters to satellite images in the water area, it is possible to obtain more accurate weather data in the area of the vessel's route. The results of calculations based on real satellite and on-board data in the waters of the Sea of Okhotsk are presented. A polynomial of the second degree is taken as a regression function. It is shown that the correction using subsatellite ship measurements more accurately characterizes the wave than the data directly received from the satellite.

PEAR: Positional-encoded Asynchronous Autoregression for satellite anomaly detection

This paper proposes a Positional-Encoded Asynchronous AutoRegression (PEAR) method for satellite anomaly detection. We empirically observe that a single classification model can hardly detect unknown anomalous situations and neglect the Markov nature of temporal satellite data. To address this, we adopt an autoregressive model to deal with the prediction of unknown anomaly for satellite data. We further propose a non-uniform temporal encoding method for asynchronous data and a median filtering method for more accurate detection. To reduce the effect of outliers, we employ an adaptive threshold selection method to achieve a more robust classification boundary. We test the proposed method on the time series prediction models including LSTM and transformers and we further employ the positional encoding strategy to improve the modeling capabilities of the Transformer model on high-frequency information. Experiments on real satellite data demonstrate that the proposed PEAR method outperforms the baseline method by 55.79%.

Orbit Predictions Using KS Elements With Earth’s Flattening

Predicting the orbit of a satellite is a fundamental requirement in many areas of aerospace, such as mission planning, satellite geodesy, re-entry prediction, maneuver planning, collision avoidance, formation flying, etc. For near-Earth orbits, the forces due to the non spherical nature of the Earth and atmospheric drag plays an important role and are mainly responsible to bring the satellite back to the Earth. Thus inclusion of the effect of these perturbing forces becomes important for precise orbit computation of near-Earth orbits. Many transformations have emerged in the literature to stabilize the equations of motion either to reduce the accumulation of local numerical errors or allowing the use of large integration step sizes, or both in the transformed space. One such transformation is known as KS transformation by Kustaanheimo and Stiefel [1], who regularized the nonlinear Kepler equation of motion and reduced it into linear differential equations of a harmonic oscillator of constant frequency. In this paper a detailed analysis is carried out for orbit prediction using KS differential equations by including the non spherical gravitational potential of the Earth as the perturbing force. Earths zonal and tesseral harmonics, which produce bands of constant deviation from the spherical field along lines of latitude and longitude, are included to precisely model the Earth’s gravitational potential. Higher order Earth’s flattening terms are included by utilizing the recurrence relations of associated Legendre polynomial and its derivatives. To know the effectiveness of the theory, the results are compared with the real IRS-1A satellite data and the results of other existing solutions for a long duration of nearly a month time. The comparison shows that the KS method provides one of the most and accurate techniques for orbit prediction available at present.

A Generalized Variable Projection Algorithm for Least Squares Problems in Atmospheric Remote Sensing

This paper presents a solution for efficiently and accurately solving separable least squares problems with multiple datasets. These problems involve determining linear parameters that are specific to each dataset while ensuring that the nonlinear parameters remain consistent across all datasets. A well-established approach for solving such problems is the variable projection algorithm introduced by Golub and LeVeque, which effectively reduces a separable problem to its nonlinear component. However, this algorithm assumes that the datasets have equal sizes and identical auxiliary model parameters. This article is motivated by a real-world remote sensing application where these assumptions do not apply. Consequently, we propose a generalized algorithm that extends the original theory to overcome these limitations. The new algorithm has been implemented and tested using both synthetic and real satellite data for atmospheric carbon dioxide retrievals. It has also been compared to conventional state-of-the-art solvers, and its advantages are thoroughly discussed. The experimental results demonstrate that the proposed algorithm significantly outperforms all other methods in terms of computation time, while maintaining comparable accuracy and stability. Hence, this novel method can have a positive impact on future applications in remote sensing and could be valuable for other scientific fitting problems with similar properties.

Fault Diagnosis for Momentum Wheels of Communication Satellite Based on Artificial Neural Network

With the advent of several fault detection techniques in modern control systems design, this paper adopted the Artificial Neural Network (ANN) Fault Detection scheme for the Fault Detection of the Attitude Control System for a Communication Satellite. In satellite applications, telemetry data can be very large, and ANN is best suited for network modeling involving large sets of data. The availability of real satellite data from Nigcomsat-1R communication satellite provided a practical platform to assess the fault detection algorithm. Results obtained showed a good correlation between raw satellite telemetry data and Neural Network model-generated results for subsequent fault detection. The fault detection models were able to detect faults, log them and provide a notification to enhance subsequent isolation and rectification. Momentum Wheel Speed and Torque were used to investigate the performance of the wheels while the Momentum Wheel Voltage and Current helped to monitor the wheel’s health state. A fault is detected if the absolute difference between original output (MW Torque) and the NN Torque output is greater than 0.012. With this, an accuracy of 100% and mean squared error of 9.8489e-6 were achieved.

Authentication for Satellite Communication Systems Using Physical Characteristics

Satellite communication networks have gained a lot of attention recently as a solution to mitigate the limitations of terrestrial networks such as stability and coverage. However, integrating satellite and terrestrial networks makes the system more vulnerable to spoofing attacks. Thus, robust and effective authentication is required. Physical layer authentication (PLA) has emerged as an alternative paradigm that uses physical characteristics to achieve authentication. In this paper, PLA is proposed for low earth orbit (LEO) satellites using the Doppler frequency shift (DS) and received power (RP) characteristics. Hypothesis testing using a threshold or machine learning (ML) is considered to discriminate between legitimate and illegitimate satellites. For ML, a one-class classification support vector machine (OCC-SVM) is employed which uses training data from only legitimate users. The performance is evaluated using real satellite data from the system tool kit (STK). Results are presented which show that the authentication rate (AR) with DS is higher than with RP at low elevation angles for both schemes, but is higher with RP at high elevation angles. Further, the ML authentication scheme provides a higher AR than the threshold scheme for a small percentage of the training data considered as outliers, but at larger percentages the OR threshold scheme is better.

Effect of Snow Cover on Detecting Spring Phenology from Satellite-Derived Vegetation Indices in Alpine Grasslands

The accurate estimation of phenological metrics from satellite data, especially the start of season (SOS), is of great significance to enhance our understanding of trends in vegetation phenology under climate change at regional or global scales. However, for regions with winter snow cover, such as the alpine grasslands on the Tibetan Plateau, the presence of snow inevitably contaminates satellite signals and introduces bias into the detection of the SOS. Despite recent progress in eliminating the effect of snow cover on SOS detection, the mechanism of how snow cover affects the satellite-derived vegetation index (VI) and the detected SOS remains unclear. This study investigated the effect of snow cover on both VI and SOS detection by combining simulation experiments and real satellite data. Five different VIs were used and compared in this study, including four structure-based (i.e., NDVI, EVI2, NDPI, NDGI) VIs and one physiological-based (i.e., NIRv) VI. Both simulation experiments and satellite data analysis revealed that the presence of snow can significantly reduce the VI values and increase the local gradient of the growth curve, allowing the SOS to be detected. The bias in the detected SOS caused by snow cover depends on the end of the snow season (ESS), snow duration parameters, and the snow-free SOS. An earlier ESS results in an earlier estimate of the SOS, a later ESS results in a later estimate of the SOS, and an ESS close to the snow-free SOS results in small bias in the detected SOS. The sensitivity of the five VIs to snow cover in SOS detection is NDPI/NDGI < NIRv < EVI2 < NDVI, which has been verified in both simulation experiments and satellite data analysis. These findings will significantly advance our research on the feedback mechanisms between vegetation, snow, and climate change for alpine ecosystems.

Secular Trends in Global Tides Derived From Satellite Radar Altimetry

AbstractPrevious studies have demonstrated that tides are subject to considerable changes on secular time scales. However, these studies rely on sea level observations from tide gauges that are predominantly located in coastal and shelf regions and therefore, the large‐scale patterns remain uncertain. Now, for the first time, satellite radar altimetry (TOPEX/Poseidon & Jason series) has been used to study worldwide linear trends in tidal harmonic constants of four major tides (M2, S2, O1, and K1). This study demonstrates both the potential and challenges of using satellite data for the quantification of such long‐term changes. Two alternative methods were implemented. In the first method, tidal harmonic constants were estimated for consecutive 4‐year periods, from which the linear change was then estimated. In the second method, the estimation of linear trends in the tidal constants of the four tides was integrated in the harmonic analysis. First, both methods were assessed by application to tide gauge data that were sub‐sampled to the sampling scheme of the satellites. Thereafter the methods were applied to the real satellite data. Results show both statistically significant decreases and increases in amplitude up to 1 mm/year and significant phase changes up to ∼0.1 deg/year. The level of agreement between altimeter‐derived trends and estimates from tide gauge data differs per region and per tide.

Deep Learning Aided Routing for Space-Air-Ground Integrated Networks Relying on Real Satellite, Flight, and Shipping Data

Current maritime communications mainly rely on satellites having meager transmission resources, hence suffering from poorer performance than modern terrestrial wireless networks. With the growth of transcontinental air traffic, the promising concept of aeronautical ad-hoc networking relying on commercial passenger airplanes is potentially capable of enhancing satellite-based maritime communications via air-to-ground and multihop air-to-air links. In this article, we conceive space-air-ground integrated networks for supporting ubiquitous maritime communications, where the low earth orbit satellite constellations, passenger air-planes, terrestrial base stations, ships, serve as the space, air, ground, and sea layer, respectively. To meet heterogeneous service requirements, and accommodate the time-varying and self-organizing nature of space-air-ground integrated networks, we propose a deep learning aided multi-objective routing algorithm, which exploits the quasi-predictable network topology and operates in a distributed manner. Our simulation results — based on real satellite, flight, and shipping data in the North Atlantic region — show that the integrated network enhances the coverage quality by reducing the end-to-end delay and by boosting the end-to-end throughput, as well as improving the path-lifetime. The results demonstrate that our deep learning aided multi-objective routing algorithm is capable of achieving near pareto-optimal performance.

An object-based sparse representation model for spatiotemporal image fusion

Many algorithms have been proposed for spatiotemporal image fusion on simulated data, yet only a few deal with spectral changes in real satellite images. An innovative spatiotemporal sparse representation (STSR) image fusion approach is introduced in this study to generate global dense high spatial and temporal resolution images from real satellite images. It aimed to minimize the data gap, especially when fine spatial resolution images are unavailable for a specific period. The proposed approach uses a set of real coarse- and fine-spatial resolution satellite images acquired simultaneously and another coarse image acquired at a different time to predict the corresponding unknown fine image. During the fusion process, pixels located between object classes with different spectral responses are more vulnerable to spectral distortion. Therefore, firstly, a rule-based fuzzy classification algorithm is used in STSR to classify input data and extract accurate edge candidates. Then, an object-based estimation of physical constraints and brightness shift between input data is utilized to construct the proposed sparse representation (SR) model that can deal with real input satellite images. Initial rules to adjust spatial covariance and equalize spectral response of object classes between input images are introduced as prior information to the model, followed by an optimization step to improve the STSR approach. The proposed method is applied to real fine Sentinel-2 and coarse Landsat-8 satellite data. The results showed that introducing objects in the fusion process improved spatial detail, especially over the edge candidates, and eliminated spectral distortion by preserving the spectral continuity of extracted objects. Experiments revealed the promising performance of the proposed object-based STSR image fusion approach based on its quantitative results, where it preserved almost 96.9% and 93.8% of the spectral detail over the smooth and urban areas, respectively.

Operational framework to predict field level crop biomass using remote sensing and data driven models

Remote Sensing (RS) based monitoring provides opportunities to acquire timely and reliable information on crop growth at diverse scales. Crop yield forecasts can help decision makers to formulate policies on maintaining national food reserves, sustaining food supply chains and attaining national food security. Inter field crop yield heterogeneity arising from varied field management and agricultural practices necessitates this forecasting to be done at field-level. Such field level information can aid farmers to identify gaps in water and farm management practices and take corrective actions, if needed. So far, no scalable tools are available to predict crop yields at field level that leverage the availability of open-access high-resolution RS data.This research implements an operational framework to predict field level crop biomass by evaluating different regression algorithms to develop data driven models, leveraging historical and near real time open-access high resolution optical satellite data from Sentinel-2, radar data from Sentinel-1 and evapotranspiration (ETa) and Net Primary Production (NPP) data from Food and Agriculture Organization’s (FAO) Water Productivity through Open-access Remotely sensed data (WaPOR) platform. NPP is used as a proxy for biomass production/yield. Five of the most commonly used regression algorithms were tested to build a data-driven model for sugarcane NPP prediction in Wonji-Shoa sugarcane estate, located in the Awash Basin, Ethiopia. The models tested were the Multivariate Linear Regression (MLR), Stepwise Multivariate Linear Regression (SMLR), Boosted Regression Trees (BRT), Support Vector Regression (SVR), and Random Forest Regression (RFR).The results revealed that for seasonal sugarcane NPP predictions, the linear regression models (MLR and SMLR) yielded more accurate predictions than the non-linear machine learning models (BRT, SVR and RFR) tested. The highest accuracy was achieved for MLR models for which estimates with 89% accuracy could be made 4 months prior to the harvest and with accuracies of 79% up to 200 days (approx. 6.5 months) before the harvest. The non-linear machine learning models, however, could not provide reliable estimates of sugarcane NPP (accuracies < 61%). Cumulative vegetation indices (VIs) were found to have higher predictive power than standard VIs for predicting future sugarcane NPP. Cumulative Enhanced Vegetation Index (EVI) was found to be the variable with the highest predictive power, followed by VH polarized Sentinel-1 Synthetic Aperture Radar data and WaPOR ETa. The study shows the usefulness of high-resolution RS information to predict seasonal NPP at field level. The methods presented here can be translated into an automated framework towards an operational system.

Satellite retrieval of benthic reflectance by combining lidar and passive high-resolution imagery: Case-I water

Under the background of global change, increasing attention has been paid to the changes of benthic habitats in shallow ocean ecosystems (e.g., seagrass beds and coral reefs). Optical satellite remote sensing via both active and passive methods plays an important role in monitoring the health of benthic habitats by retrieving benthic reflectance spectra, but it remains difficult to accurately retrieve benthic reflectance spectra from only active or passive remote sensing because of the coupling between water column scattering and benthic reflectance. Here, we developed a semi-analytical model to retrieve benthic reflectance spectra in Case-I waters by combining active lidar and passive high-resolution imagery. Based on two-stream radiative transfer theory, the analytical relationship among the remote sensing reflectance (Rrs) and water column reflectance (Rw), benthic reflectance (Rb), diffuse attenuation coefficient (Kd), and water depth was established. The lidar data at a certain wavelength were applied to derive the water depth and the chlorophyll concentration (chl) along the lidar track. Then, the values of Kd at different wavelengths were estimated from the derived chl. In addition, we established a look-up-table (LUT) for the relationship between Rw and chl and water depth using Hydrolight simulation, and the Rw values at different wavelengths were then estimated by the lidar-derived chl and water depth. Finally, Rb(λ) values at different wavelengths along the lidar track were retrieved from the Rrs(λ) values observed by passive high-resolution imagery and the values of Rw(λ), Kd(λ), and water depth derived by lidar observation. The accuracy of the model was verified by using the Hydrolight simulated datasets, and the high correlation coefficient (R) revealed promising model performance for different benthic habitats, e.g., R > 0.9 for typical clean shallow water (chl = 0.5 mg/m3, H < 4 m) for the wavelength range of 400–640 nm. The model was further applied to real satellite data from ICESat-2 lidar and passive high-resolution satellite imagery (multispectral imagery of Sentinel-2 and hyperspectral imagery of Zhuhai−1) at two different benthic habitat sites (seagrass beds in Xincun Bay and coral reefs in the Huaguang Reef) in the South China Sea, and the results revealed that the model could reproduce the benthic spectra in both magnitude and shape. Overall, the proposed model can reliably yield benthic reflectance spectra along the lidar track without any requirement on prior knowledge, which should be beneficial for further benthic habitat health monitoring.

Surface reflectance and pXRF for assessing soil weathering indexes

The use of satellite images is a widespread technique for estimating soil attributes. However, what is the potential of satellite data on the prediction of soil weathering indices? How much in their performance is affected by spectral mixing and data resolution? There is no consensus for answering these questions. Hence, the main objectives of this work were: i) to show the difference of spectral signature between prepared and field-condition samples; ii) to compare the convoluted data from laboratory spectroradiometer with the spectral signatures from Sentinel-2 and Landsat-8 satellites; and iii) to build and evaluate predictive models for soil weathering indices using diffuse reflectance spectroscopy from samples under natural conditions, air-dried-sieved and from satellite images. Soil samples were collected from the first 1-cm depth in 27 locations along seven catenas located in a 163-ha area in São Paulo State, Brazil. The samples were analyzed by a Vis-NIR-SWIR spectrometer without sample preparation (field conditions) and after being air-dried and sieved (prepared). The air-dried-sieved samples were also analyzed by an X-ray fluorescence equipment. Reflectance data were obtained from Sentinel and Landsat for the same 27 locations. The spectral data from the lab were convoluted to respective bands of the real satellite data. Predictive models for soil weathering indices using the convoluted laboratory data and the satellite data were also built and compared. The reflectance of samples under field conditions contrasted significantly between soils derived from basalt and sedimentary rocks. In general, the reflectance intensity of prepared samples was 80% higher than that observed for field-conditions, which can be explained by the roughness and the structure effect. However, some cases showed an opposite trend, with a smaller relative reflectance from the air-dried-sieved samples. These trends support the hypothesis of aggregates heterogeneity and clay occlusion. The sample preparation can increase or decrease the reflectance of the soil surface material, depending, mainly, on roughness, soil structure and particles occlusion by the structure. The Landsat spectral signatures were more similar to those from convoluted laboratory than to the Sentinel ones. The satellite performance on the estimation of the soil weathering indices is comparable to those from convoluted laboratory data, but it is limited by spectral resolution. Sentinel and Landsat achieved ratio of performance to interquartile distance (RPIQ) from 1.53 to 2.38 and from 1.56 to 2.42 on soil weathering indices estimation, respectively. The models for Plagioclase Index of Alteration (PIA) and Si/Ti [Si and (Ti and Al) ratio] indices were considered the best ones (RPIQ from 2.17 to 3.6; and RPIQ from 1.90 to 5.56; respectively) for both laboratory and satellite data; in turn, Si/Al, Si (Al + Fe) and Chemical Index of Alteration (CIA) indices performed poorly/reasonably (RPIQ <1.92).

Effects of Mixture Mode on the Canopy Bidirectional Reflectance of Coniferous–Broadleaved Mixed Plantations

One of the main initiatives for China to achieve the goal of being carbon neutral before 2060 is transforming monocultures into mixed plantations in subtropical China, because mixed forests possess a higher quality than monocultures in various ways. Very high spatial resolution (VHR) satellite imagery is very promising to precisely monitor the transformation process under the premise of clarifying the canopy reflectance anisotropy of mixed plantations. However, it is almost impossible to understand the canopy reflectance anisotropy of mixed plantations with real satellite data due to the extreme lack of multiangular VHR satellite images. In this study, the effects of the mixture mode on the canopy bidirectional reflectance factor (BRF) were comprehensively analyzed with simulated VHR images. The three-dimensional (3D) Discrete Anisotropic Radiative Transfer model (DART) was used to construct a pure coniferous scene, a pure broadleaved scene, and 27 coniferous–broadleaved mixed plantation scenes containing 3 mixture patterns (i.e., mixed by single trees, mixed by stripes, and mixed by patches) and 9 mixing proportions (i.e., from 10% to 90% with the interval of 10%), and to simulate red (R) and near-infrared (NIR) VHR images for these 3D scenes at both the solar principal plane (SPP) and perpendicular plane (PP) under different solar-viewing geometries. Negative correlations were generally found between the canopy BRF and the ratio of conifers in a mixed stand. The anisotropy of conifer dominated plantations is more prominent than broadleaf dominated plantations, especially for the single tree mixture. Although the level of anisotropy is much lower for PP than SPP, it should not be ignored, especially for the R band. Observations under large viewing zenith angles at PP are more preferred to study the effect of mixing proportions, followed by forward observations at SPP. The R band image has higher potential to distinguish mixture patterns for broadleaf-dominated situations, while the NIR band image has a higher potential for conifer-dominated situations. Furthermore, the canopy BRF generally increases with the solar zenith angle, and one meter can be considered as the optimal spatial resolution for the optical monitoring of the mixture mode. The findings of the current study add some valuable theoretical knowledge for the accurate monitoring of coniferous–broadleaved mixed plantations with VHR imagery.

Combined Adjustment Pipeline for Improved Global Geopositioning Accuracy of Optical Satellite Imagery With the Aid of SAR and GLAS

Owing to the widespread availability of multiple remote sensing methods, regions are often covered by satellite observation from multiple sources. However, the geometric positioning accuracy of optical satellites, which are the most common data source, is generally insufficient. An intuitive idea to improve accuracy is to integrate the dominant accuracy of multi-source information. Previous studies have used synthetic aperture radar (SAR) and Geoscience Laser Altimetry System (GLAS) data with improved geo-positioning accuracy. However, the accuracy and applicability of existing combination methods are unsatisfactory because of various restrictions. In this study, a pipeline with automatic extraction of tie points and combined adjustment was designed based on non-stereo SAR images and GLAS data, considering both the high precision and wide distribution of multi-source data. Experiments using real satellite data (ZY-3, GF-7, GF-3, and ICESat-2) and test areas covering complex landforms demonstrated the feasibility and effectiveness of the proposed pipeline. The geo-positioning accuracy in three experimental areas reached 3.16, 3.36, and 3.17 m in the horizontal direction and 1.45, 1.25, and 1.28 m in the vertical direction. Our pipeline not only permits high accuracy close to the nominal accuracy of heterogeneous reference data but can also be extended to global high-precision geo-positioning without ground control.

RFI Source Localization Based on Joint Sparse Recovery in Microwave Interferometric Radiometry

Recently, the radio frequency interference (RFI) poses a growing threat to the Microwave Interference Radiometer with Aperture Synthesis (MIRAS) led by the European Space Agency, whose scientific goal is to monitor the soil moisture and ocean salinity of the Earth. RFI localization is a critical step to mitigate the impact of RFI sources on the brightness temperature (BT) maps. In this paper, we propose an effective and robust RFI source localization approach combined with the joint sparse recovery (JSR) theory by using multi-snapshot data. The JSR method utilizes the joint sparsity property of RFI sources in the spatial domain to improve the robustness and accuracy for RFI source localization problem. First, we propose a JSR model by exploiting the joint sparsity of multi-snapshot visibility data with RFI contained from MIRAS after conducting the field of view (FOV) registration. Then, we present a greedy algorithm using the orthogonal matching pursuit on multi-snapshot data (MSOMP) to localize RFI sources. Results on synthetic data and real satellite data both show that, compared with the previous existing approaches, the proposed method has a better performance on the mitigation of the localization accuracy bias, especially for a low BT value range case and a mixed BT value range case with multiple RFI sources.

A modified fuzzy clustering approach in unsupervised classification for detecting the mixed pixels of satellite images

The major problem of remote sensing images is mixed pixels, available in the data which degrades the quality, accuracy of the image classification and object recognition. To overcome the problem of mixed pixel in a real satellite data a modified K -means clustering algorithm and a modified fuzzy C -means clustering algorithm, are discussed. The algorithms are developed by modifying the membership function of the standard K -means clustering algorithm (FKM) and the standard fuzzy C -means algorithm (FCM). The performance of the proposed algorithms is discussed and compared with the traditional fuzzy K -means algorithm and the traditional FCM algorithm. Results on classification and segmentation of satellite images reveal that the suggestive algorithms are robust and effective.