Retrieval Algorithm Research Articles

A Framework for Iterative Phase Retrieval Technique Integration into Atmospheric Adaptive Optics—Part I: Wavefront Sensing in Strong Scintillations

The objective of this study, which is divided into two parts, is twofold: to address long-standing challenges in the sensing of atmospheric turbulence-induced wavefront aberrations under strong scintillation conditions via a comparative analysis of several basic scintillation-resistant wavefront sensing (SR-WFS) architectures and iterative phase retrieval (IPR) techniques (Part I, this paper), and to develop a framework for the potential integration of SR-WFS techniques into practical closed-loop non-astronomical atmospheric adaptive optics (AO) systems (Part II). In this paper, we consider basic SR-WFS mathematical models and phase retrieval algorithms, tradeoffs in sensor design and phase retrieval technique implementation, and methodologies for WFS parameter optimization and performance assessment. The analysis is based on wave-optics numerical simulations imitating realistic turbulence-induced phase aberrations and intensity scintillations, as well as optical field propagation inside the SR-WFSs. Several potential issues important for the practical implementation of SR-WFS and IPR techniques, such as the requirements for phase retrieval computational grid resolution, tolerance with respect to optical element misalignments, and the impact of camera noise and input light non-monochromaticity, are also considered. The results demonstrate that major wavefront sensing requirements desirable for AO operation under strong intensity scintillations can potentially be achieved by transitioning to novel SR-WFS architectures, based on iterative phase retrieval techniques.

A novel AMSR2 retrieval algorithm for global C-band vegetation optical depth and soil moisture (AMSR2 IB): Parameters' calibration, evaluation and inter-comparison

Effective monitoring of soil and vegetation properties on a global scale is essential for better understanding climate changes, hydrological dynamics, and ecological processes. Passive microwave remote sensing at C-band radio frequency, with long observation period and relatively high penetration capability, has been widely used to retrieve soil moisture (SM) and vegetation optical depth (C-VOD). The retrieval process is generally achieved by inversion of the τ-ω radiative transfer model, which depends on crucial parameters such as effective scattering albedo (ω) and soil roughness (HR) for accurate retrievals. Current SM/C-VOD retrieval algorithms, such as the Land Parameter Retrieval Model (LPRM), predominantly rely on globally-constant ω and HR values, ignoring the inherent sensitivity of those parameters to varying soil conditions and vegetation types. To evaluate the impact of ω and HR variables on SM and C-VOD retrievals and to improve their accuracy, this study proposed and evaluated a novel retrieval approach from AMSR2 C-band observations during 2017–2020 using the C-band Microwave Emission of the Biosphere (C-MEB) model. We evaluated two new retrieval algorithms, considering either a globally-constant calibration or a land cover-based calibration of ω and HR. As a benchmark for the calibration, we optimized the values of ω and HR by evaluating the retrieved SM against in situ measurements from the International Soil Moisture Network (ISMN) and OzNet hydrological monitoring networks. The main originality compared to previous algorithms is that i) it includes a comprehensive calibration exploring the optimal values of ω and HR, applicable globally or tailored to specific land cover; ii) field SM measurements were leveraged to constrain the calibrated value of ω and HR.For the globally-constant calibration, the optimal values of ω = 0.05 and HR = 0.1 were found to yield the best results. For the land cover-based calibration, an inverse relationship between ω/HR and canopy height was observed, with ω ranging from 0.04 to 0.06 and HR ranging from 0.1 to 0.7 for heights between 0 and 30 m. The algorithm employing a land cover-based calibration (INRAE Bordeaux 2, IB2) exhibited better performance than the one utilizing a globally-constant calibration (INRAE Bordeaux 1, IB1) in evaluating retrieved SM against in situ measurements, as well as in evaluating C-VOD vs various vegetation variables including aboveground biomass (AGB), tree cover, canopy height and several optical vegetation indices. Comparison with LPRM suggested that our IB2 C-VOD retrievals present improved performances in terms of both spatial and temporal results with all considered vegetation variables (spatial correlation (R) between various vegetation variables and C-VOD of 0.76–0.83 for IB2 vs 0.69–0.79 for LPRM), and exhibited lower saturation effects when compared with AGB. In addition, the IB2 SM produced lower root mean squared error (RMSE) (0.147 vs 0.217 m3/m3), bias (−0.03 vs 0.09 m3/m3), and ubRMSE (0.066 vs 0.067 m3/m3) when compared with in situ measurements, although it showed a lower R compared to LPRM SM.

Comparative Evaluation of Pre-Trained Language Models for Biomedical Information Retrieval.

Finding relevant information in the biomedical literature increasingly depends on efficient information retrieval (IR) algorithms. Cross-Encoders, SentenceBERT, and ColBERT are algorithms based on pre-trained language models that use nuanced but computable vector representations of search queries and documents for IR applications. Here we investigate how well these vectorization algorithms estimate relevance labels of biomedical documents for search queries using the OHSUMED dataset. For our evaluation, we compared computed scores to provided labels by using boxplots and Spearman's rank correlations. According to these metrics, we found that Sentence-BERT moderately outperformed the alternative vectorization algorithms and that additional fine-tuning based on a subset of OHSUMED labels yielded little additional benefit. Future research might aim to develop a larger dedicated dataset in order to optimize such methods more systematically, and to evaluate the corresponding functions in IR tools with end-users.

Thermal vibrations in the inversion of dynamical electron scattering

Relativistic forward scattering of electrons at finite temperature involves the incoherent superposition of diffraction patterns formed by different snapshots of thermal atomic displacements. In experiments, thermal vibrations lead to thermal diffuse scattering (TDS), partly dominating diffraction patterns of thick specimens. This study sheds light on the effects of TDS on solutions to the inverse scattering problem using combined real- and diffraction-space information acquired in a scanning transmission electron microscope (STEM) to retrieve the object's phase. Using frozen phonon multislice within the Einstein approximation, realistic ground truth data of 20-nm-thick SrTiO3 is generated and subjected to contemporary inverse multislice schemes to retrieve the projected Coulomb potential slicewise. We first classify phase retrieval algorithms as to their assumptions on periodicity along the incident beam direction, as well as pixelwise and parametrized reconstruction methods. It is found that pixelwise object reconstructions are capable of retrieving structural details qualitatively while being prone to contain TDS-related artifacts which can result in unphysical potentials. For pixelwise reconstructions of multiple independent specimen slices, we observe that the origin of TDS, i.e., thermal atomic displacements, starts to emerge naturally. However, the quantitative assessment tends to too small mean squared thermal displacements, also when reconstructing multiple object modes. Using an atomistically parametrized inversion strategy which exploits the explicit separation of thermal vibrations and potentials, temperature and chemistry of the specimen can be retrieved quantitatively with high accuracy. Published by the American Physical Society 2024

Bridging spatio-temporal discontinuities in global soil moisture mapping by coupling physics in deep learning

The launch of Soil Moisture Active Passive (SMAP) satellite in 2015 has resulted in significant achievements in global soil moisture mapping. Nonetheless, spatiotemporal discontinuities in the soil moisture products have arisen due to the limitations of its orbit scanning gap and retrieval algorithms. To address these issues, this paper presents a physics-constrained gap-filling method, PhyFill for short. The PhyFill method employs a partial convolutional neural network technique to explore spatial domain features of the original SMAP soil moisture data. Then, it incorporates variations in soil moisture induced by precipitation events and dry-down events as penalty terms in the loss function, thereby accounting for monotonicity and boundary constraints in the physical processes governing the dynamic fluctuations of soil moisture. The PhyFill model was applied to SMAP soil moisture data, resulting in continuous spatially daily soil moisture data on a global scale. Three validation strategies are employed: visual inspection through global pattern, simulated missing-region validation, and soil moisture validation with in situ measurements. The results indicated that the reconstructed soil moisture achieved a higher spatial coverage with satisfactory spatial continuity with neighbouring pixels. The simulated validation of the missing regions revealed that the averaged unbiased root mean square difference (ubRMSD) and correlation coefficient (R) were 0.01 m3/m3 and 0.99, respectively versus the gap filled SMAP product. The core validation sites demonstrated that the reconstructed soil moisture data has a consistent ubRMSD compared with the original SMAP soil moisture data (0.04 m3/m3vs. 0.04 m3/m3). The PhyFill method can generate globally continuous, high accurate soil moisture estimates, providing remarkable support for advanced hydrological applications, e.g., global soil moisture dry-down events and patterns.

X-ray dark-field via spectral propagation-based imaging

Dark-field X-ray imaging visualizes scattering from unresolved microstructure. Most dark-field imaging techniques rely on crystals or structured illumination, but recent work has shown that dark-field effects are observable in straightforward propagation-based imaging (PBI). Based on the single-material X-ray Fokker–Planck equation with an a priori dark-field energy dependence, we propose an algorithm to extract phase and dark-field effects from dual-energy PBI images. We successfully apply the dark-field retrieval algorithm to simulated and experimental dual-energy data, and show that by accounting for dark-field effects, projected thickness reconstruction is improved compared to the classic Paganin algorithm. With the emergence of spectral detectors, the method could enable single-exposure dark-field imaging of dynamic and living samples.

Accurate two-step random phase retrieval approach without pre-filtering based on hyper ellipse fitting

In this work, we propose a hyper ellipse fitting-based high-precision random two-frame phase shifting algorithm to improve the accuracy of phase retrieval. This method includes a process of Gram-Schmidt orthonormalization, followed by a hyper ellipse fitting procedure. The Gram-Schmidt orthonormalization algorithm constructs a quadrature fringe pattern relative to the original fringe pattern. These two quadrature fringe patterns are then fed into the hyper ellipse fitting procedure, which reconstructs the phase map and refines the background light to produce the final accurate phase of interest. Due to the hyper ellipse fitting procedure, the demodulation results are significantly improved in many cases. This method allows us to design a two-shot phase reconstruction algorithm without the need for least squares iteration or pre-filtering, effectively mitigating residual background to the greatest extent. It combines the advantages of both the Gram-Schmidt orthonormalization method and the Lissajous ellipse fitting method, making our hyper ellipse fitting approach a simple, flexible, and accurate phase retrieval algorithm. Experiments show that by using the weighted least squares method and adjusting the weights, this method can prioritize data points with more significant information or higher reliability, ensuring more accurate estimation of the ellipse parameters.

LangchainIQ: Intelligent Content and Query Processing

Purpose: The purpose of this research is to introduce and evaluate a comprehensive framework called langchain(component of Large Language Model), designed to optimize data analysis and visualization processes across various business domains. The framework integrates advanced computational techniques with user-friendly interfaces to meet the growing demand for efficient information processing tools in research and industry settings. Design/Methodology/Approach: The framework consists of three primary components: PDF answering, CSV analytics, and data visualization using the LIDA library. Integration of advanced technologies such as the Mistral 7B model for language processing, Faiss for similarity search, and the LIDA library for data visualization. Detailed implementation steps include content processing, embedding using OpenAI embeddings, storage and retrieval using Faiss, and query handling using Mistral 7B. This involves breaking down PDF and CSV content into chunks, embedding them, and utilizing advanced algorithms for efficient data retrieval and visualization. Findings/Result: The fine-tuned Mistral 7B model significantly enhances data extraction speed compared to traditional models like Llama. Users can effectively query and extract specific information from PDFs and CSVs using natural language, facilitated by advanced AI models. The LIDA library automates the generation of insightful visualizations from processed data, enhancing data interpretation and decision-making. Originality/Value: Introducing langchain as a versatile framework that addresses the complexities of data analysis and visualization and it’s use in business analysis. Paper Type: Technical Research.

Development of L-band fully polarimetric SAR algorithm for forest biomass retrieval using 7SD and random forest regression

Forest biomass is essential for carbon budgeting, biodiversity health, and climate change research. This study developed a novel L-band fully polarimetric SAR method for estimating aboveground biomass (AGB) in a forested region in India. Biomass was estimated using fully polarimetric ALOS-2/PALSAR-2 data sets and field AGB measurements. To estimate forest AGB, the algorithm employs a seven-component scattering power decomposition (7SD) and a random forest regression (RFR) machine-learning approach. By implementing the 7SD model, the seven scattering powers were extracted from ALOS-2/PALSAR-2 inverted for aboveground biomass. The field data were used to validate the biomass estimated by the model. The 7SD model estimated AGB in Shivamogga was consistent with field measured biomass. The root mean squared error (RMSE) and relative RMSE with respect to mean AGB were 21.94 Mg/ha and 19.46, respectively, which are within acceptable ranges. The 7SD model was also used for cross-validation at Tundi Forest, where the relative RMSE with respect to the mean AGB was 22.9. The scattering powers generated from L-band fully polarimetric data can be useful in tropical forest AGB estimation.

Machine learning-based retrieval of total column water vapor over land using GMI-sensed passive microwave measurements

ABSTRACT The Global Precipitation Measurement (GPM) Microwave Imager (GMI) is a microwave (MW) radiometer that has near-global coverage and frequent revisit time. To date, operational total column water vapor (TCWV) data records from the GPM GMI sensor have been exclusively offered over oceanic regions. It is challenging to retrieve TCWV over land from satellite MW measurements because of varying land surface characteristics. In this paper, a novel Light Gradient Boosting Machine-based retrieval algorithm is proposed to derive TCWV over land from GMI-sensed MW brightness temperature (BT) observations. The GMI-observed MW BT at 18.7 GHz and 23.8 GHz, differential BT between 18.7 GHz and 23.8 GHz, latitude, longitude, and month are selected and utilized as the input variables of the retrieval approach, because of their strong correlation with satellite-sensed MW TCWV retrievals. Instead of surface emissivity data or radiative transfer model, we take into account the spatial and temporal elements, namely latitude, longitude, and month. The training of the retrieval method is performed based on ground-based TCWV estimates from worldwide 4,471 Global Navigation Satellite System (GNSS) stations in 2017. The performance of the newly proposed retrieval algorithm is independently validated in a worldwide coverage using reference TCWV from additional 4,341 GNSS stations in 2018–2020 and 605 radiosonde stations in 2017–2020. The newly retrieved TCWV estimates over land have a correlation coefficient of 0.76 and 0.83, a root-mean-square error (RMSE) of 5.82 mm and 6.02 mm, a relative RMSE of 34.91% and 34.36%, and a mean bias of 0.02 mm and −0.42 mm compared to reference TCWV from GNSS and radiosonde, respectively. The performance of the retrieval algorithm is satisfactory when compared to that of land-purpose TCWV of other satellite missions, though we have not used either surface emissivity data or radiative transfer model. This result increases confidence in retrieving TCWV over land from satellite-sensed MW BT measurements based on machine learning using ground-based TCWV observations. The newly developed retrieval algorithm has the potential for integration into operational satellite missions or meteorological services, thereby enhancing weather forecasting, climate modeling, and other relevant applications.

Machine learning based aerosol and ocean color joint retrieval algorithm for multiangle polarimeters over coastal waters

NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, recently launched in February 2024, carries two multiangle polarimeters (MAPs): the UMBC Hyper-Angular Rainbow Polarimeter (HARP2) and SRON Spectropolarimeter for Planetary Exploration One (SPEXone). Measurements from these MAPs will greatly advance ocean ecosystem and aerosol studies as their measurements contain rich information on the microphysical properties of aerosols and hydrosols. The Multi-Angular Polarimetric Ocean coLor (MAPOL) joint retrieval algorithm has been developed to retrieve aerosol and ocean color information, which uses a vector radiative transfer (RT) model as the forward model. The RT model is computationally expensive, which makes processing a large amount of data challenging. FastMAPOL was developed to expedite retrieval using neural networks to replace the RT forward models. As a prototype study, FastMAPOL was initially limited to open ocean applications where the ocean Inherent Optical Properties (IOPs) were parameterized in terms of one parameter: chlorophyll-a concentration (Chla). In this study we further expand the FastMAPOL joint retrieval algorithm to incorporate NN based forward models for coastal waters, which use multi-parameter bio-optical models. In addition, aerosols are represented by six components, i.e., fine mode non absorbing insoluble (FNAI), brown carbon (BrC), black carbon (BC), fine mode non absorbing soluble (FNAS), sea salt (SS) and non-spherical dust (Dust). Sea salt and dust are coarse mode aerosols, while the other components are fine mode. The sizes and spectral refractive indices are fixed for each aerosol component, while their abundances are retrievable. The multi-parameter bio-optical model and aerosol components are chosen to represent the coastal marine environment. The retrieval algorithm is applied to synthetic measurements in three different configurations of MAPs in the PACE mission: HARP2 observations only, SPEXone observations only and combined HARP2 and SPEXone observations. The retrieval results from synthetic measurements show that for aerosol retrieval the SPEXone-only configuration works equally well with the HAPR2-only configuration. On the other hand, for ocean color retrieval the SPEXone instrument provides better information due to its larger spectral coverage. For the surface parameters (wind speed), HARP2 measurements provide better information due to its wide field of view. Combined measurement configuration HARP2+SPEXone performed the best to retrieve all aerosol, ocean color, and surface parameters. We also studied the impact of sun glint to aerosol and ocean color retrievals. The retrieval test revealed that wind speed and absorbing aerosol retrieval improves significantly when including measurements at glint geometries. Furthermore, the retrieval algorithm is equipped with modules for atmospheric correction and bidirectional reflectance distribution (BRDF) correction to obtain the remote sensing reflectance, which enables ocean biogeochemistry studies using the PACE polarimeter data.

Using convolutional neural network denoising to reduce ambiguity in X-ray coherent diffraction imaging.

The inherent ambiguity in reconstructed images from coherent diffraction imaging (CDI) poses an intrinsic challenge, as images derived from the same dataset under varying initial conditions often display inconsistencies. This study introduces a method that employs the Noise2Noise approach combined with neural networks to effectively mitigate these ambiguities. We applied this methodology to hundreds of ambiguous reconstructed images retrieved from a single diffraction pattern using a conventional retrieval algorithm. Our results demonstrate that ambiguous features in these reconstructions are effectively treated as inter-reconstruction noise and are significantly reduced. The post-Noise2Noise treated images closely approximate the average and singular value decomposition analysis of various reconstructions, providing consistent and reliable reconstructions.

Secure and efficient data storage with Rivest Shamir Adleman algorithm in cloud environment

Cloud computing rapidly is a prerequisite and releases resources with minimal management effort. The surfacing of the cloud has significantly distorted the general insight into infrastructure, software services, and development models. In contrast to single-key encryption models based on public or private keys (PKs), hybrid encryption systems combine encryption methods using symmetric or asymmetric methods. Various hybrid algorithms fail to meet users’ expectations regarding data security and cannot prevent all security risks. The secure and efficient data storage and retrieval (SEDSR) algorithm was developed for scalable key management between the content owner, cloud user, and service providers in an un-trusted cloud environment. In the implementation, the SEDSR combines the Rivest Shamir Adleman (RSA) algorithm 4096 key length with a primary symmetric key method to provide adequate and compact security with optimal retrieval systems in the cloud. Based on the experimental evaluation, the SEDSR minimizes 1.7 seconds of encryption times (ET) and 1.5 seconds of decryption time (DT) and improves by 34% throughput (TRP) compared to existing parameters.

Methodology and Uncertainty Analysis of Methane Flux Measurement for Small Sources Based on Unmanned Aerial Vehicles

The top–down emission rate retrieval algorithm (TERRA) method for calculating the net flux out of a box has been employed by other researchers to assess large sources of methane release. This usually requires a manned aircraft drone with powerful performance to fly over the boundary layer. Few studies have focused on low-altitude box sampling mass balance methods for small sources of methane release, such as at maximum flight altitudes of less than 100 m. The accuracy and sources of uncertainty in such a method still need to be determined as they differ from the conditions of large sources. Nineteen flights were conducted to detect methane emissions from Chinese oil field well sites using a measurement system consisting of a quadcopter and methane, wind speed, wind direction, air pressure, and temperature sensors. The accuracy and uncertainty of the method are discussed. The average absolute relative error of the measurement is 18.5%, with an average uncertainty of 55.75%. The uncertainty is mainly caused by the wind speed and direction, and the background CH4 concentration. The main paths to reduce uncertainty and improve accuracy for low-altitude box sampling include subtracting the background concentration during flux retrieval, enhancing the accuracy of methane measurements, selecting a period of downwind dominant or wind direction change of less than 30 degrees, and ensuring a maximum flight height greater than 50 m with a horizontal distance from the pollution source center of less than 75 m. The results show that TERRA-based low-altitude box sampling is suitable for quantifying methane release rates from small sources.

Temperature Is Likely an Important Omission in Interpreting Vegetation Optical Depth

AbstractVegetation optical depth (VOD) satellite microwave retrievals provide significant insights into vegetation water content and responses to hydroclimatic changes. While VOD variations are commonly linked to dry biomass and live fuel moisture content (LFMC), the impact of canopy temperature (Tc) remains overlooked in large‐scale studies. Here, we investigated the impact of Tc on L‐band (1.4 GHz) and X‐band (10.7 GHz) VOD at diurnal and seasonal timescales. Synthetic benchmark VOD was created using realistic fields of Tc, LFMC, and biomass in an electromagnetic model. Perturbation experiments revealed that Tc strongly affects diurnal VOD variations at both L‐band and X‐band. Seasonally, while biomass emerges as the largest contributor to VOD variations in 70% (at X‐band) and 90% (at L‐band) of our study region, Tc and LFMC still play substantial roles. The findings stress the importance of refining retrieval algorithms to distinguish Tc, LFMC, and biomass effects for future VOD applications in ecohydrology.

Inferences of water–ice cloud physics determined from MAVEN/IUVS aerosol retrievals and Mars GCM comparisons

We performed retrievals of dust and water–ice optical depths at select times spanning Mars years (MYs) 33 through 36. We used data taken by the Imaging Ultraviolet Spectrograph (IUVS) instrument aboard the Mars Atmosphere and Volatile Evolution (MAVEN) spacecraft, whose precessing orbit allowed IUVS to image the diurnal evolution of water–ice clouds at local times throughout this MY range where they have not been studied. We created a radiative-transfer retrieval algorithm and used it to simultaneously fit the dust and water–ice optical depth of each usable spectrum in IUVS apoapse data for the majority of the MAVEN mission. We compared our results to other datasets to ensure our retrievals produced plausible values and discuss potential reasons for discrepancies. We obtained global climate model (GCM) simulations from two state-of-the-art models and compared our results to both sets of simulations. We show that there are significant differences between our water–ice retrievals and both sets of simulations, both in the spatial cloud patterns and column-integrated optical depths. We discuss the different approaches both models used for simulating these clouds and the implications for accurately modeling these clouds in future GCM work.

Mathematical analysis of human motion vision capture image processing based on artificial intelligence

In order to understand the mathematical analysis of human motion vision capture image processing, a mathematical analysis of human motion vision capture image processing based on artificial intelligence is proposed. This paper firstly introduces the research progress, classification and several commonly used vision-based human motion tracking methods of motion capture technology. Secondly, the process and framework of capturing human motion video by using ordinary cameras and marking nodes are proposed, and the automatic tracking algorithm based on Camshift and Kalman filter is adopted. It verifies the effectiveness of the system, changes the traditional way of motion capture, and makes the process of capture more convenient when the capture effect meets the requirements. Finally, the performance of the human motion capture data retrieval algorithm based on video is evaluated comprehensively. It is compared with the latest literature in this field. In terms of time efficiency, for each online retrieval of data set, the proposed algorithm takes 0.056 s, while the methods of other scholars take an average of 1.5 s. Meanwhile, experiments are also conducted on public databases, proving the universality and scalability of the proposed algorithm. The algorithm proposed in this paper has greater advantages than the most advanced method of the same type, which verifies the effectiveness of the algorithm proposed in this paper.

Sea surface barometry with an O2 differential absorption radar: retrieval algorithm development and simulation

Sea surface air pressure observations are a significant gap in the current Earth observing systems. This study addresses retrieval algorithm development and the evaluation of the potential impact of instrumental and environmental uncertainties on sea level pressure retrievals for the measurements of O2 differential absorption radar systems operating at three spectrally evenly spaced close-frequency bands (65.5, 67.75, and 70.0 GHz). A simulated northern hemispheric summer case is used to simulate retrieval uncertainties. To avoid high attenuation and a low signal-to-noise ratio, radar measurements from weather conditions with a rain rate ≥1 mm/h are not used in the retrieval. This study finds that a retrieval algorithm combining all three channels, i.e., the 3-channel approach, can effectively mitigate major atmospheric and sea surface influences on sea surface air pressure retrieval. The major uncertainty of sea surface pressure retrieval is due to the standard deviation in radar power returns. Analysis and simulation demonstrate the potential of global sea surface pressure observations with errors of about 1∼2 mb, which is urgently needed for the improvement of numerical weather prediction models. Future work will emphasize instrument development and field experiments. It is anticipated that an O2 differential absorption radar system will be available for meteorological applications in a few years.

A novel laser interferometric method for thin film thickness measurement in gas–liquid intermittent flow

Abstract In horizontal intermittent flow, the long bubbles move toward the center of the pipe due to inertia, forming the thin liquid film above the long bubbles. Accurate measurement of the liquid film thickness is crucial for heat and mass transfer. In this paper, laser interferometric technology is innovatively introduced to measure the film thickness of the intermittent flow, and the thin liquid film is detected with a resolution of 100 nm. Considering the curvature of the circular pipe wall, which leads to divergent reflected light, the effect of the pipe wall on the interference pattern is explored by the ray tracing technique. A two-dimensional interpolation phase retrieval algorithm based on light intensity is proposed to reconstruct the thickness of the liquid film, and the average error is less than 1.86%. Benefiting from the exceptionally high resolution, research is conducted on the thin liquid film at the top of the horizontal intermittent flow, revealing its dependence on the sub-regimes and gas–liquid velocities.

First atmospheric aerosol-monitoring results from the Geostationary Environment Monitoring Spectrometer (GEMS) over Asia

Abstract. Aerosol optical properties have been provided by the Geostationary Environment Monitoring Spectrometer (GEMS), the world's first geostationary-Earth-orbit (GEO) satellite instrument designed for air quality monitoring. This study describes improvements made to the GEMS aerosol retrieval (AERAOD) algorithm, including spectral binning, surface reflectance estimation, cloud masking, and post-processing, along with validation results. These enhancements aim to provide more accurate and reliable aerosol-monitoring results for Asia. The adoption of spectral binning in the lookup table (LUT) approach reduces random errors and enhances the stability of satellite measurements. In addition, we introduced a new high-resolution database for surface reflectance estimation based on the minimum-reflectance method, which was adapted to the GEMS pixel resolution. Monthly background aerosol optical depth (BAOD) values were used to estimate hourly GEMS surface reflectance consistently. Advanced cloud-removal techniques have been implemented to significantly improve the effectiveness of cloud detection and enhance aerosol retrieval quality. An innovative post-processing correction method based on machine learning has been introduced to address artificial diurnal biases in aerosol optical depth (AOD) observations. In this study, we investigated selected aerosol events, highlighting the capability of GEMS in monitoring and providing insights into hourly aerosol optical properties during various atmospheric events. The performance of the GEMS AERAOD products was validated against the Aerosol Robotic Network (AERONET) and Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) data for the period from November 2021 to October 2022. GEMS AOD at 443 nm demonstrated a strong correlation with AERONET AOD at 443 nm (R = 0.792). However, it exhibited biased patterns, including the underestimation of high AOD values and overestimation of low-AOD conditions. Different aerosol types (highly absorbing fine aerosols, dust aerosols, and non-absorbing aerosols) exhibited distinct validation results. The retrievals of GEMS single-scattering albedo (SSA) at 443 nm agreed well with the AERONET SSA at 440 nm within reasonable error ranges, with variations observed among aerosol types. For GEMS AOD at 443 nm exceeding 0.4 (1.0), 42.76 % (56.61 %) and 67.25 % (85.70 %) of GEMS SSA data points fell within the ±0.03 and ±0.05 error bounds, respectively. Model-enforced post-processing correction improved GEMS AOD and SSA performance, thereby reducing the diurnal variation in the biases. The validation of the retrievals of GEMS aerosol layer height (ALH) against the CALIOP data demonstrates good agreement, with a mean bias of −0.225 km and 55.29 % (71.70 %) of data points falling within ±1 km (1.5 km).