Saturation Adjustment Research Articles

Improving supercooled liquid water representation in the microphysical scheme ICE3

AbstractMost numerical weather prediction (NWP) models have a significant bias in predicting supercooled liquid water (SLW). For this reason, icing risk diagnostic tools do not use supercooled liquid water forecast by the models as an input parameter, but rather temperature and humidity, which are forecast better than SLW. The main objective of this study is to improve the SLW representation in the microphysical scheme ICE3 (Three Ice categories) used in the operational Applications de la Recherche à l'Opérationnel à Méso‐Echelle (AROME) model. For this purpose, several parametrizations of the microphysical processes were evaluated to find a better representation of SLW in the Mesoscale Non‐Hydrostatic model (MESO‐NH), which also uses ICE3. Elements of the microphysical scheme of the HARMONIE‐AROME model and work carried out by the Centre National de Recherches Météorologiques (CNRM) have been tested and compared with the current scheme. After a preliminary study, three parametrizations of the microphysical scheme were selected, in which the processes of ice initiation, snow and graupel collection of cloud droplets, condensation, Bergeron–Findeisen, and saturation adjustment were modified. Then, MESO‐NH simulations were performed and compared with observations from the In‐Cloud ICing and Large‐drop Experiment (ICICLE) airborne campaign. Three case studies were used with different icing weather conditions such as freezing rain, freezing drizzle, lake effect, etc. The results show a better representation of SLW with a greater presence of cloud droplets for colder temperatures up to 30 C. However, the liquid water content remains underestimated and the ice mass is overestimated. The ice initiation and cloud droplet collection by snow and graupel play a major role in the SLW representation. Parametrizations with restrictive ice initiation criteria reduce cloud droplet consumption and provide better agreement with observations. The results are promising and need to be investigated further with more cases and in the operational model AROME.

A modified hue and range preserving color assignment function with a component-wise saturation adjustment for color image enhancement

This paper presents a new approach to enhancing color images by modifying an affine color assignment function. This function maps colors to pixels of the enhanced gray-scale image in a way that improves the visual quality of the image, particularly in darker regions. The main goal of our method is to finely adjust saturation, correct saturation loss in specific image regions, and preserve the original image’s range and hue. Our proposed method follows a two-step process. First, it enhances the intensity image using a combination of global and local histogram equalization methods. This results in an overall improved appearance by redistributing pixel intensities and enhancing contrast. Then, modified color mapping functions are applied to assign colors to each pixel of the enhanced gray-scale image. The aim is to adjust saturation by amplifying the maximally saturated color image. Additionally, we introduce two new color-weighted maps to evaluate pixel importance from the maximally saturated image. This contributes to saturation control in the final enhanced image. Compared to alternative color mapping algorithms, our model preserves the original color of pixels in challenging areas and fine-tunes saturation based on parameter settings.

Importance of CCN activation for fog forecasting and its representation in the two‐moment microphysical scheme LIMA

AbstractThe work presented in this article studies the impact of cloud condensation nuclei (CCN) activation for fog forecasting and improves its parameterization in the LIMA (Liquid, Ice, Multiple Aerosols) two‐moment microphysical scheme, building upon the Local And Non‐local Fog EXperiment (LANFEX) field campaign observations, specifically the intensive observation period (IOP) 1 and the DEMISTIFY intercomparison. Large‐eddy simulations were performed with the Meso‐NH model, first using a prognostic supersaturation allowing us to compute the number of activated CCN at each time step, and then with the usual saturation adjustment hypothesis and a diagnostic maximum supersaturation. The prognostic supersaturation method provided very good results, similar to those from earlier simulations of this case using a bin scheme, and was thus used as a reference. In contrast, the diagnostic maximum supersaturation method strongly overestimated droplet numbers and produced a too‐thick fog. Thus, improvements to the maximum supersaturation diagnostic were proposed, by (1) revising the temperature tendency and (2) accounting for pre‐existing cloud droplets in the activation parameterization. These improvements resulted in a simulation in good agreement with observations and the reference simulation, and are promising for use in numerical weather prediction systems with a lower resolution and/or a longer time step.

Towards a more reliable forecast of ice supersaturation: concept of a one-moment ice-cloud scheme that avoids saturation adjustment

Abstract. A significant share of aviation's climate impact is due to persistent contrails. Thus, avoiding the creation of contrails that exert a warming impact is a crucial step in approaching the goal of sustainable air transportation. For this purpose, a reliable forecast of when and where persistent contrails are expected to form is needed (i.e. a reliable prediction of ice supersaturation). With such a forecast at hand, it would be possible to plan aircraft routes on which the formation of persistent contrails can be avoided. One problem on the way to these forecasts is the current systematic underestimation of the frequency and degree of ice supersaturation at cruise altitudes in numerical weather prediction due to the practice of “saturation adjustment”. In this common parameterisation, the air inside cirrus clouds is assumed to be exactly at ice saturation, while measurement studies have found cirrus clouds to be quite often out of equilibrium. In this study, we propose a new ice-cloud scheme that overcomes saturation adjustment by explicitly modelling the decay of the in-cloud humidity after nucleation, thereby allowing for both in-cloud super- and subsaturation. To achieve this, we introduce the in-cloud humidity as a new prognostic variable and derive the humidity distribution in newly generated cloud parts from a stochastic box model that divides a model grid box into a large number of air parcels and treats them individually. The new scheme is then tested against a parameterisation that uses saturation adjustment, where the stochastic box model serves as a benchmark. It is shown that saturation adjustment underestimates humidity, both shortly after nucleation, when the actual cloud is still highly supersaturated, and also in aged cirrus if the temperature keeps decreasing, as the actual cloud remains in a slightly supersaturated state in this case. The new parameterisation, on the other hand, closely follows the behaviour of the stochastic box model in any considered case. The improvement in comparison with saturation adjustment is largest if slow updraughts occur in relatively clean air in models with a high spatial and temporal resolution. We conclude that our parameterisation is promising but needs further testing in more realistic frameworks.

Meta underwater camera: A smart protocol for underwater image enhancement

Underwater images captured by an underwater camera normally suffer from visual degradation issues, such as color deviations, low contrasts, and detail blurs. Existing studies tend to address these issues separately by individual techniques, such that underwater image visibility can hardly be improved in an overall, consistent manner. To address this limitation, we present a smart protocol for underwater image enhancement. The protocol comprises a comprehensive cascade of seven image enhancement techniques, i.e., attenuated channel compensation, white balance, tone mapping, hue-saturation-lightness (HSL) model-based saturation adjustment, contrast stretching, gamma correction, and high-pass fusion. The protocol is smartly configured by reinforcement learning. Specifically, a set of underwater image multi-color space features are considered as a state, a set of parameter values for the protocol as an action, and an increment of a non-reference visual preference score as a reward. By the reinforcement learning strategy, the parameter values for the seven techniques in the protocol are optimally configured as a whole, resulting in optimal underwater image enhancement results. We refer to the functionality of the smart protocol as a meta underwater camera (MUC) methodology because it operates behind an underwater camera’s operation of capturing underwater images but provides underwater images of high visibility beyond those of low visibility originally obtained from the underwater camera. Qualitative and quantitative empirical results validate that our MUC methodology outperforms state-of-the-art underwater image enhancement methods. We release the reproducible code at https://gitee.com/wanghaoupc/MUC_UnderwaterImageEnhancement for public evaluations.

Three-Dimensional Static Instability of Gravity Waves and a Possible Parameterization of the Associated Wave Breaking

Abstract Parameterizations of subgrid-scale gravity waves (GWs) in atmospheric models commonly involve the description of the dissipation of GWs. Where they dissipate, GWs have an increased effect on the large-scale flow. Instabilities that trigger wave breaking are an important starting point for the route to dissipation. Possible destabilizing mechanisms are numerous, but the classical vertical static instability is still regarded as a key indicator for the disposition to wave breaking. In this work, we investigate how the horizontal variations associated with a GW could alter the criterion for static instability. To this end, we use an extension of the common parcel displacement method. This three-dimensional static stability analysis predicts a significantly larger range of instability than does the vertical static stability analysis. In this case, the Lindzen-type saturation adjustment to a state of marginal stability is perhaps a less suitable ansatz for the parameterization of the GW breaking. To develop a possible ansatz for the GW dissipation due to three-dimensional instability, we apply the methods of irreversible thermodynamics, which are embedded in the Gibbs formalism of dynamics. In this way, the parameterization does not only satisfy the second law of thermodynamics, but it can also be made consistent with the conservation of energy and further (non-)conservation principles. We develop the parameterization for a discrete spectrum of GW packets. Offline computations of GW drag and dissipative heating rates are performed for two vertical profiles of zonal wind and temperature for summer and winter conditions from CIRA data. The results are compared to benchmarks from the literature.

Understanding Controlling Factors of Extratropical Humidity and Clouds with an Idealized General Circulation Model

Abstract This paper examines the physical controls of extratropical humidity and clouds by isolating the effects of cloud physics factors in an idealized model. The Held–Suarez dynamical core is used with the addition of passive water vapor and cloud tracers, allowing cloud processes to be explored cleanly. Separate saturation adjustment and full cloud scheme controls are used to consider the strength of advection–condensation theory. Three sets of perturbations to the cloud scheme are designed to test the model’s sensitivity to the physics of condensation, sedimentation, and precipitation formation. The condensation and sedimentation perturbations isolate two key differences between the control cases. First, the sub-grid-scale relative humidity distribution assumed for the cloud macrophysics influences the location and magnitude of the extratropical cloud maxima, which interrupt the isentropic transport of moisture to the polar troposphere. Second, within the model’s explicit treatment of cloud microphysics, re-evaporation of hydrometeors moistens and increases clouds in the lower troposphere. In contrast, microphysical processes of precipitation formation (specifically, the ratio of accretion to autoconversion) have negligible effects on humidity, cloudiness, and precipitation apart from the strength of the large-scale condensation and formation cycle. In addition, counterintuitive relationships—such as cloud condensate and cloud fraction responding in opposing directions—emphasize the need for careful dissection of physical mechanisms. In keeping with advection–condensation theory, circulation sets the patterns of humidity, clouds, and precipitation to first order, with factors explored herein providing secondary controls. The results substantiate the utility of such idealized modeling and highlight key cloud processes to constrain.

Warm-phase spectral-bin microphysics in ICON: reasons of sensitivity to aerosols

The warm-phase version of the Spectral-Bin Microphysics (SBM) scheme has been implemented into the ICON model and has been tested in a supercell storm simulation. The results of two-moment (2M) bulk scheme which exists in ICON and SBM were compared in all aspects related to warm processes, especially with regards to sensitivity to aerosols.In all simulations a strong storm with maximum updrafts exceeding 25 m/s was simulated. The maximum vertical velocity in the 2M bulk scheme does not depend on aerosols. However, the maximum vertical velocity in SBM with high aerosol concentration was found substantially higher than in the case of low aerosol concentration.Despite the insensitivity of maximum vertical velocity to aerosols, the microphysics of the 2M scheme is sensitive to aerosols. Both SBM and 2M schemes showed faster rain formation in case of low aerosol concentration; both schemes indicate that first raindrops form when the mean volume radius exceeds ~15 μm. In both schemes the first raindrops form near cloud top in an undiluted core and fall along the cloud edges at the mature stage. At the decaying stage, raindrops fall through the cloud core. The zone of rainfall at the lower few kilometers coincides with the zones of strong downdrafts. A substantial effect of aerosols on intensity and spatial distribution of precipitation was found. In the case of low aerosol concentration raindrops form earlier and at lower altitudes. As a result, the area covered by rain is lower, but the maximum of the rain intensity is higher than in case of high aerosol concentration. Finally, we discuss the hindering effects of large supersaturation via mass loading on the cloud development in case of low aerosol concentration, and examine the validity of the well-known saturation adjustment procedure.

Deep unsupervised endoscopic image enhancement based on multi-image fusion

Background and objectiveA deep unsupervised endoscopic image enhancement method is proposed based on multi-image fusion to achieve high quality endoscope images from poorly illuminated, low contrast and color deviated images through an unsupervised mapping and deep learning network without the need for ground truth. MethodsFirstly, three image enhancement methods are used to process original endoscopic images to obtain three derived images, which are then transformed into HSI color space. Secondly, a deep unsupervised multi-image fusion network (DerivedFuse) is proposed to extract and fuse features of the derived images accurately by utilizing a new no-reference quality metric as loss function. I-channel images of the three derived images are inputted into the DerivedFuse network to enhance the intensity component of the original image. Finally, a saturation adjustment function is proposed to adaptive adjusting the saturation component of HSI color space to enrich the color information of the original input image. ResultsThree evaluation metrics: Entropy, Contrast Improvement Index (CII) and Average Gradient (AG) are used to evaluate the performance of the proposed method. The results are compared with that of fourteen state-of-the-art algorithms. Experiments on endoscopic image enhancement show that the Entropy value of our method is 3.27% higher than the optimal entropy value of comparison algorithms. The CII of our proposed method is 6.19% higher than that of comparison algorithms. The AG of our method is 7.83% higher than the optimal AG of comparison algorithms. ConclusionsThe proposed deep unsupervised multi-image fusion method can obtain image information details, enhance endoscopic images with high contrast, rich and natural color information, visual and image quality. Sixteen doctors and medical students have given their assessments on the proposed method for assisting clinical diagnoses.

The FastEddy® Resident‐GPU Accelerated Large‐Eddy Simulation Framework: Moist Dynamics Extension, Validation and Sensitivities of Modeling Non‐Precipitating Shallow Cumulus Clouds

AbstractHerein we describe the moist dynamics formulation implemented within the graphics processing unit‐resident large‐eddy simulation FastEddy® model, which includes a simple saturation adjustment scheme for condensation and evaporation processes. Two LES model intercomparison exercises for non‐precipitating shallow cumulus clouds are simulated in order to validate this model extension, including a static forcing and a time‐dependent forcing case. Overall, we find our dynamical, thermodynamical and microphysical quantities, along with turbulence variability and fluxes, to be commensurate with the corresponding model intercomparison results. In addition, sensitivities to specific model settings are investigated. Among these settings, it is shown that boundary layer and cloud layer structure and characteristics are sensitive to use of higher‐order advection schemes impacting the vertical distribution of cloud content and associated turbulence statistics. Increasing the timescale of the saturation scheme leads to enhanced liquid water presence and decreases vertical velocity variance within the cloud deck. In some cases, these sensitivities agree with the model‐to‐model variability reported in the intercomparison exercises, highlighting the important role of specific model implementation choices in the context of shallow cumulus convection simulations. These analyses and findings also provide the basis for future extensions and applications of FastEddy® for modeling moist convection and precipitation scenarios.

White-box content camouflage attacks against deep learning

Deep learning has achieved remarkable success in a wide range of computer vision tasks. However, recent researches suggest that deep learning systems are vulnerable to a variety of attacks. Security concerns have been raised regarding the training or inference phase of deep learning models in the last few years, and the research field about the vulnerability of the pre-processing components in these models is still developing. In this paper, we systematically examine white-box content camouflage attacks on five types of pre-processing modules in deep learning systems: scaling, sharpening, Gamma correction, contrast adjustment, and saturation adjustment. We assume that an attacker's goal is to generate camouflage examples that show inconsistent visual semantics before and after pre-processing. Under the white-box setting (where the pre-processing algorithms and their parameters are known), we formulate content camouflage attacks as an optimization problem in which perceptual losses in the source and target images are smoothly calculated by a multi-scale discriminator to improve the camouflaging effect of the attack example. We evaluate our content camouflage attacks by conducting a series of experiments on two example groups as well as two real-world datasets, i.e., CIFAR-10 and FER-2013. The experimental results show that with good camouflaging ability, our attacks are effective against deep learning systems, and outperform prevalent scaling camouflage attacks by generating examples with better quality and a higher attack success rate. The proposed camouflage attacks are also extended to the four commonly used pre-processing algorithms, and yield good results. Furthermore, we discuss the effect of varying the parameters of several image pre-processing algorithms under our attacks and analyze' the reasons for their vulnerability.

A Real-Time Semi-Supervised Deep Tone Mapping Network

Tone mapping operators (TMOs) can compress the range of high dynamic range (HDR) images so that they can be displayed normally on the low dynamic range (LDR) devices. Recent TMOs based on deep neural networks can produce impressive results, but there are still some shortcomings. On the one hand, their supervised learning procedure requires a high-quality paired dataset which is hard to be accessed. On the other hand, they are too slow and heavy to meet the needs of practical applications. This paper proposes a real-time deep semi-supervised learning TMO to solve the above problems. The proposed method learns in a semi-supervised manner by combining the adversarial loss, cycle consistency loss, and the pixel-wise loss. The first two can simulate the image distributions in the real world from the unpaired LDR data and the latter can learn the guidance of paired LDR labels. In this way, the proposed method only requires HDR sources, unpaired high-quality LDR images, and a few well tone-mapped HDR-LDR pairs as training data. Furthermore, the proposed method divides tone mapping into luminance mapping and saturation adjustment and then processes them simultaneously. By this strategy, we can reconstruct each component more precisely. Based on the aforementioned improvements, we propose a lightweight tone mapping network that is efficient in tone mapping task (up to 5000x parameters-saving and 27x time-saving compared to the learning-based TMOs). Both quantitative and qualitative results demonstrate that the proposed method performs favorable against state-of-the-art TMOs.

Haze Elimination Model-Based Color Saturation Adjustment With Contrast Correction

Haze Elimination Model-Based Color Saturation Adjustment With Contrast Correction

Impacts of long-range-transported mineral dust on summertime convective cloud and precipitation: a case study over the Taiwan region

Abstract. As one of the most abundant atmospheric aerosols and effective ice nuclei, mineral dust affects clouds and precipitation in the Earth system. Here numerical experiments are carried out to investigate the impacts of dust aerosols on summertime convective clouds and precipitation over the mountainous region of Taiwan by acting as ice-nucleating particles. We run the Weather Research and Forecasting model (WRF) with the Morrison two-moment and spectral-bin microphysics (SBM) schemes at 3 km resolution, using dust number concentrations from a global chemical transport model (GEOS-Chem-APM). The case study indicates that the long-range-transported mineral dust, with relatively low number concentrations, can notably affect the properties of convective clouds (ice and liquid water contents, cloud top height, and cloud coverage) and precipitation (spatial pattern and intensity). The effects of dust are evident during strong convective periods, with significantly increased ice water contents in the mixed-phase regime via the enhanced heterogeneous freezing. With both the Morrison and SBM schemes, we see the invigoration effects of dust aerosols on the convective intensity through enhanced condensation and deposition latent heating. The low-altitude dust particles are uplifted to the freezing level by updrafts, which, in turn, enhance the convective cloud development through immersion freezing and convective invigoration. Compared to the Morrison scheme, the SBM scheme predicts more realistic precipitation and different invigoration effects of dust. The differences are partially attributed to the saturation adjustment approach utilized in the bulk scheme, which leads to a stronger enhancement of condensation at midlatitudes to low altitudes and a weaker deposition increase at the upper level.

Revisiting adiabatic fraction estimations in cumulus clouds: high-resolution simulations with a passive tracer

Abstract. The process of mixing in warm convective clouds and its effects on microphysics are crucial for an accurate description of cloud fields, weather, and climate. Still, they remain open questions in the field of cloud physics. Adiabatic regions in the cloud could be considered non-mixed areas and therefore serve as an important reference to mixing. For this reason, the adiabatic fraction (AF) is an important parameter that estimates the mixing level in the cloud in a simple way. Here, we test different methods of AF calculations using high-resolution (10 m) simulations of isolated warm cumulus clouds. The calculated AFs are compared with a normalized concentration of a passive tracer, which is a measure of dilution by mixing. This comparison enables the examination of how well the AF parameter can determine mixing effects and the estimation of the accuracy of different approaches used to calculate it. Comparison of three different methods to derive AF, with the passive tracer, shows that one method is much more robust than the others. Moreover, this method's equation structure also allows for the isolation of different assumptions that are often practiced when calculating AF such as vertical profiles, cloud-base height, and the linearity of AF with height. The use of a detailed spectral bin microphysics scheme allows an accurate description of the supersaturation field and demonstrates that the accuracy of the saturation adjustment assumption depends on aerosol concentration, leading to an underestimation of AF in pristine environments.

A Numerical Analysis of Six Physics‐Dynamics Coupling Schemes for Atmospheric Models

AbstractSix strategies to couple the dynamical core with physical parameterizations in atmospheric models are analyzed from a numerical perspective. Thanks to a suitably designed theoretical framework featuring a high level of abstraction, the truncation error analysis and the linear stability study are carried out under weak assumptions. Indeed, second‐order conditions are derived which are not influenced either by the specific formulation of the governing equations, nor by the number of parameterizations, nor by the structural design and implementation details of the time‐stepping methods. The theoretical findings are verified on two idealized test beds. Particularly, a hydrostatic model in isentropic coordinates is used for vertical slice simulations of a moist airflow past an isolated mountain. Self‐convergence tests show that the sensitivity of the prognostic variables to the coupling scheme may vary. For those variables (e.g., momentum) whose evolution is mainly driven by the dry dynamics, the truncation error associated with the dynamical core dominates and hides the error due to the coupling. In contrast, the coupling error of moist variables (e.g., the precipitation rate) emerges gradually as the spatio‐temporal resolution increases. Eventually, each coupling scheme tends toward the formal order of accuracy, upon a careful treatment of the grid cell condensation. Indeed, the well‐established saturation adjustment may cap the convergence rate to first order. A prognostic formulation of the condensation and evaporation process is derived from first principles. This solution is shown effective to alleviate the convergence issues in our experiments. Potential implications for a complete forecasting system are discussed.

Supersaturation, buoyancy, and deep convection dynamics

Abstract. Motivated by recent discussions concerning differences of convective dynamics in polluted and pristine environments, the so-called convective invigoration in particular, this paper provides an analysis of factors affecting convective updraft buoyancy, such as the in-cloud supersaturation, condensate and precipitation loading, and entrainment. We use the deep convective period from simulations of daytime convection development over land discussed in our previous publications. An entraining parcel framework is used in the theoretical analysis. We show that for the specific case considered here, finite (positive) supersaturation noticeably reduces pseudo-adiabatic parcel buoyancy and cumulative convective available potential energy (cCAPE) in the lower troposphere. This comes from keeping a small fraction of the water vapor in a supersaturated state and thus reducing the latent heating. Such a lower-tropospheric impact is comparable to the effects of condensate loading and entrainment in the idealized parcel framework. For the entire tropospheric depth, loading and entrainment have a much more significant impact on the total CAPE. For the cloud model results, we compare ensemble simulations applying either a bulk microphysics scheme with saturation adjustment or a more comprehensive double-moment scheme with supersaturation prediction. We compare deep convective updraft velocities, buoyancies, and supersaturations from all ensembles. In agreement with the parcel analysis, the saturation-adjustment scheme provides noticeably stronger updrafts in the lower troposphere. For the simulations predicting supersaturation, there are small differences between pristine and polluted conditions below the freezing level that are difficult to explain by standard analysis of the in-cloud buoyancy components. By applying the piggybacking technique, we show that the lower-tropospheric buoyancy differences between pristine and polluted simulations come from a combination of temperature (i.e., latent heating) and condensate loading differences that work together to make polluted buoyancies and updraft velocities slightly larger when compared to their pristine analogues. Overall, the effects are rather small and contradict previous claims of a significant invigoration of deep convection in polluted environments.

Integrated cloud macro- and micro-physics schemes with kinetic treatment of condensation processes for global models

A new parameterization scheme was developed to remove the saturation adjustment assumption and resolve the condensation process in the grid-scale cloud macrophysics scheme to build an integrated cloud microphysics scheme for global climate models. By applying a saturation prediction equation with calculations based on cloud hydrometeor properties, supersaturation or subsaturation can be determined within the macrophysics scheme. This treatment provides the basis for condensation calculation and allows the Wegener–Bergeron–Findeisen process to be resolved explicitly to render a realistic liquid–ice partition in mixed-phase clouds. The cloud fraction scheme was modified based on physics principles to complement the condensation scheme. The new scheme's performance was examined by incorporating it into the Community Atmosphere Model version 5 (CAM5) single-column model to simulate a Tropical Warm Pool–International Cloud Experiment (TWP–ICE) case. The results revealed that grid-scale cloud properties are sensitive to the condensation process's treatment, and the new scheme can produce a more reasonable cloud fraction and liquid–ice partition than the original CAM5. The theory-based scheme developed in this study may provide insight for addressing consistency between the macrophysical and microphysical schemes in global climate models.

Comments on “Do Ultrafine Cloud Condensation Nuclei Invigorate Deep Convection?”

AbstractHere we elaborate on the deficiencies associated with the theoretical arguments and model simulations in a paper by Grabowski and Morrison (2020, hereafter GM20) that argued convective invigoration by aerosols does not exist. We show that the invigoration can be supported by both accurate theoretical analysis and explicit physics modeling with prognostic supersaturation and aerosols. Negligible invigoration by aerosols via drop freezing in GM20 was explained by a complete compensation between the heating effect from the freezing of extra liquid water and the extra loading effect during droplet ascending. But the reality is that droplet ascending then freezing occur at different locations and time scales, producing complex nonlinear responses that depend on the duration and location of the forcing. Also, this argument neglects the effect of off-loading of precipitating ice particles, increases in condensation during ascending, and riming and deposition accompanying droplet freezing. Regarding the warm-phase invigoration, the quasi-steady assumption for supersaturation as adopted in GM20 makes condensation independent of droplet number and size, therefore an incorrect interpretation of warm-phase invigoration. We illustrate that the quasi-steady assumption is invalid for updrafts of deep convective clouds in clean conditions because of the high acceleration of vertical velocity and the fast depletion of droplets by raindrop formation and accretion. Any assumption imposed on supersaturation, such as quasi-steady approximation and saturation adjustment, leads to errors in the evaluation of aerosol effects on diffusional growth and related buoyancy. Furthermore, we demonstrate that the piggybacking approach they used cannot prove or disprove the convective invigoration.

A Fast Image Dehazing Algorithm for Highway Tunnel Based on Artificial Multi-exposure Image Fusion

Single image dehazing has been a challenging problem due to its ill-posed nature. This paper proposed a fast image dehazing algorithm for highway tunnel based on artificial multi-exposure image fusion. The original blurred images are transformed into under-exposed images using a series of gamma-correction operations. Meanwhile, a multi-exposure image fusion algorithm based on two-scale decomposition is proposed. It exploits the guided filter to separate each multi-exposure image into base and detail layers. The fused base and detail layers are integrated into the fused images. Then, a linear saturation adjustment method is utilized to enhance the saturation in spatial domain. Through the above image processing steps, the final haze-free images can be obtained. In this paper, the superiority of the proposed image dehazing algorithm based on artificial multi-exposure image fusion is demonstrated through theoretical analysis and experimental verification. It can quickly improve the visibility of foggy images, and make the processed images colorful and clear.