Mesoscale Phenomena Research Articles

The sugar cube: Network control and emergence in stereoediting reactions.

Stereochemical editing strategies have recently enabled the transformation of readily accessible substrates into rare and valuable products. Typically, site selectivity is achieved by minimizing kinetic complexity by using protecting groups to suppress reactivity at undesired sites (substrate control) or by using catalysts with tailored shapes to drive reactivity at the desired site (catalyst control). We propose "network control," a contrasting paradigm that exploits hidden interactions between rate constants to greatly amplify modest intrinsic biases and enable precise multisite editing. When network control is applied to the photochemical isomerization of hexoses, six of the eight possible diastereomers can be selectively obtained. The amplification effect can be viewed as a mesoscale phenomenon between the limiting regimes of kinetic control in simple chemical systems and metabolic regulation in complex biological systems.

On Some Limitations of Current Machine Learning Weather Prediction Models

AbstractMachine Learning (ML) is having a profound impact in the domain of Weather and Climate Prediction. A recent development in this area has been the emergence of fully data‐driven ML prediction models which routinely claim superior performance to that of traditional physics‐based models. We examine some aspects of the forecasts produced by three of the leading current ML models, Pangu‐Weather, FourCastNet and GraphCast, with a focus on their fidelity and physical consistency. The main conclusion is that these ML models are not able to properly reproduce sub‐synoptic and mesoscale weather phenomena and lack the fidelity and physical consistency of physics‐based models and this has impacts on the interpretation of their forecasts and their perceived skill. Balancing forecast skill and physical realism will be an important consideration for future ML models.

The multi-scale coupled model: a new framework capturing wind farm–atmosphere interaction and global blockage effects

Abstract. The growth in the number and size of wind energy projects in the last decade has revealed structural limitations in the current approach adopted by the wind industry to assess potential wind farm sites. These limitations are the result of neglecting the mutual interaction of large wind farms and the thermally stratified atmospheric boundary layer. While currently available analytical models are sufficiently accurate to conduct site assessments for isolated rotors or small wind turbine clusters, the wind farm's interaction with the atmosphere cannot be neglected for large-size arrays. Specifically, the wind farm displaces the boundary layer vertically, triggering atmospheric gravity waves that induce large-scale horizontal pressure gradients. These perturbations in pressure alter the velocity field at the turbine locations, ultimately affecting global wind farm power production. The implication of such dynamics can also produce an extended blockage region upstream of the first turbines and a favorable pressure gradient inside the wind farm. In this paper, we present the multi-scale coupled (MSC) model, a novel approach that allows the simultaneous prediction of micro-scale effects occurring at the wind turbine scale, such as individual wake interactions and rotor induction, and meso-scale phenomena occurring at the wind farm scale and larger, such as atmospheric gravity waves. This is achieved by evaluating wake models on a spatially heterogeneous background velocity field obtained from a reduced-order meso-scale model. Verification of the MSC model is performed against two large-eddy simulations (LESs) with similar average inflow velocity profiles and a different capping inversion strength, so that two distinct interfacial gravity wave regimes are produced, i.e. subcritical and supercritical. Interfacial waves can produce high blockage in the first case, as they are allowed to propagate upstream. On the other hand, in the supercritical regime their propagation speed is less than their advection velocity, and upstream blockage is only operated by internal waves. The MSC model not only proves to successfully capture both local induction and global blockage effects in the two analyzed regimes, but also captures the interaction between the wind farm and gravity waves, underestimating wind farm power by about only 2 % compared with the LES results. Conversely, wake models alone cannot distinguish between differences in thermal stratification, even if combined with a local induction model. Specifically, they are affected by a first-row overprediction bias that leads to an overestimation of the wind farm power by 13 % to 20 % for the modeled regimes.

Improving Weather Forecasts for Sailing Events Using a Combination of a Numerical Forecast Model and Machine Learning Postprocessing

Accurate predictions of wind and other weather phenomena are essential for making informed strategic and tactical decisions in sailing. Sailors worldwide utilize current state-of-the-art forecasts, yet such forecasts are often insufficient because they do not offer the high temporal and geographic resolution required by sailors. This paper examines wind forecasting in competitive sailing and demonstrates that traditional wind forecasts can be improved for sailing events by using an integration of traditional numerical modeling and machine learning (ML) methods. Our primary objective is to provide practical and more precise wind forecasts that will give sailors a competitive edge. As a case study, we demonstrate the capabilities of our proposed methods to improve wind forecasting at Lake Kinneret, a popular sailing site. The lake wind pattern is highly influenced by the area’s topographic features and is characterized by unique local and mesoscale phenomena at different times of the day. In this research, we simulate the Kinneret wind during the summers of 2015–2021 in up to one-kilometer resolution using the Weather Research and Forecasting (WRF) atmospheric model. The results are used as input for convolutional neural network (CNN) and multilayer perceptron (MLP) ML models to postprocess and improve the WRF model accuracy. These advanced ML models are trained using training datasets based on the WRF data as well as real data measured by the meteorological service, and subsequently, a validation process of the trained ML model is performed on unseen datasets against site-specific meteorological service observations. Through our experimental analysis, we demonstrate the limitations of the WRF model. It uncovers notable biases in wind direction and velocity, particularly a persistent northern bias in direction and an overestimation of wind strength. Despite its inherent limitations, this study demonstrates that the integration of ML models can potentially improve wind forecasting due to the remarkable prediction accuracy rate achieved by the CNN model, surpassing 95%, while achieving partial success for the MLP model. Furthermore, a successful CNN-based preliminary forecast was effectively generated, suggesting its potential contribution to the future development of a user-friendly tool for sailors.

A Review of Remote Sensing of Atmospheric Profiles and Cloud Properties by Ground-Based Microwave Radiometers in Central China

Thermodynamic and liquid water profiles can be retrieved by a ground-based microwave radiometer (MWR) in nearly all weather conditions, which is useful for detecting mesoscale phenomena. This paper reviews the advances in remote sensing of atmospheric profiles and cloud properties by MWR in central China. Comparative studies indicate that MWR retrieval accuracy is different under various skies, especially those that decay under precipitation. The off-zenith method is proven to be capable of reducing the impact of precipitation and snow on MWR retrieval accuracy. Application studies demonstrate that MWR retrievals are helpful for early warning of rainstorms, hailstorms, and thunderstorms. Moreover, MWR retrievals provide a way to study cloud properties. The temporal variations of cloud occurrence frequency (COF) and liquid water path (LWP) are different for low, middle, and high clouds, and the vertical distribution of COF is also different in autumn and other seasons. Note that MWR can infer valid retrievals over the eastern Tibetan Plateau due to the weak precipitation over there. Also, cloud properties over the eastern Tibetan Plateau present differences from those over central China, and this is related to the different characteristics of atmospheric water vapor between these two regions. To bring more benefits for mechanism study and early warning of severe weather and numerical weather prediction, the decayed accuracy of MWR zenith retrievals under precipitation should be resolved. And combining MWR with other instruments is necessary for MWR application in detecting multi-layer clouds and ice clouds.

Acoustic tomographic inversion of 3D temperature fields with mesoscale anomaly in the South China Sea

Acoustic tomographic inversion is based on travel times measured along the transmission paths between all station pairs to reconstruct three-dimensional temperature structures with mesoscale anomalies. In this study, tomographic simulation experiments were designed based on the Hybrid Coordinate Ocean Model (HYCOM) reanalysis data to reconstruct mesoscale phenomena from travel time data obtained from five, seven, and nine stations in the South China Sea over a domain of 100 × 100 km. The travel times for each station pair were calculated in the vertical section using the Bellhop acoustic ray simulation method. Six Empirical orthogonal function (EOF) modes of sound speed along the sound transmission paths in a vertical slice were used to formulate the inversion equations. The horizontal-slice distributions of temperature in the tomography domain were reconstructed using the grid-segmented method for each depth layer. For station-to-station distances greater than 100 km, the performance of inversion was best for the seven-station case rather than for the nine-station case, with the highest horizontal resolution of the three cases. This case study concluded that the seven-station case rather than the nine-station case provided an optimal station number for reconstructing the three-dimensional temperature fields.

A shear stress parametrization for arbitrary wind farms in conventionally neutral boundary layers

In the context of large off-shore wind farms, power production is influenced greatly by the turbine array's interaction with the atmospheric boundary layer. One of the most influencing manifestations of such complex interaction is the increased level of shear stress observed within the farm. This leads to higher momentum fluxes that affect the wind speed at the turbine locations and in the cluster wake. At the wind farm entrance, an internal boundary layer (IBL) grows due to the change in effective roughness imposed by the wind turbines, and for large enough clusters, this can reach the unperturbed boundary layer height in what is referred to as the fully developed regime. Downwind, a second IBL starts growing, while the shear stress profile decays exponentially to its unperturbed state. In the present study, we propose a simple analytical model for the vertical profile of the horizontal shear stress components in the three regions identified above. The model builds upon the top-down model of Meneveau (J. Turbul., vol. 13, 2012, N7), and assumes that the flow develops in a conventionally neutral boundary layer. The proposed parametrization is verified successfully against large-eddy simulations, demonstrating its ability to capture the vertical profile of horizontal shear stress, and its evolution both inside and downwind of the wind farm. Our findings suggest that the developed model can prove extremely useful to enhance the physical grounds on which new classes of coupled wind farm engineering models are based, leading to a better estimation of meso-scale phenomena affecting the power production of large turbine arrays.

Dynamic Nanocrystal Superlattices with Thermally Triggerable Lubricating Ligands.

The size-dependent and collective physical properties of nanocrystals (NCs) and their self-assembled superlattices (SLs) enable the study of mesoscale phenomena and the design of metamaterials for a broad range of applications. However, the limited mobility of NC building blocks in dried NCSLs often hampers the potential for employing postdeposition methods to produce high-quality NCSLs. In this study, we present tailored promesogenic ligands that exhibit a lubricating property akin to thermotropic liquid crystals. The lubricating ability of ligands is thermally triggerable, allowing the dry solid NC aggregates deposited on the substrates with poor ordering to be transformed into NCSLs with high crystallinity and preferred orientations. The interplay between the dynamic behavior of NCSLs and the molecular structure of the ligands is elucidated through a comprehensive analysis of their lubricating efficacy using both experimental and simulation approaches. Coarse-grained molecular dynamic modeling suggests that a shielding layer from mesogens prevents the interdigitation of ligand tails, facilitating the sliding between outer shells and consequently enhancing the mobility of NC building blocks. The dynamic organization of NCSLs can also be triggered with high spatial resolution by laser illumination. The principles, kinetics, and utility of lubricating ligands could be generalized to unlock stimuli-responsive metamaterials from NCSLs and contribute to the fabrication of NCSLs.

High deformation and delamination mechanisms explained through mesoscale and nanoscale phenomena in pearlitic steel wires

Pearlitic steel wires subjected to severe cold drawing deformation exhibit a unique combination of high strength and excellent ductility, making them suitable for critical applications such as hanging bridge cables, crane cables and tire cord. Torsion tests up to rupture are commonly used in the industry for quality control purposes. A flat fracture surface is indicative of wire aptitude, while a delaminated surface suggests degradation of mechanical properties. This paper aims to expand the understanding of the structural evolution and deformation mechanisms that occur during wire drawing, as well as the origin of delamination problems. Different deformation and strengthening mechanisms are identified and discussed, applying different microscopy techniques (optical microscopy, scanning electron microscopy, electron backscatter diffraction), synchrotron diffraction and atom probe tomography. The information is correlated with the cementite stability on the structure studied at meso, nano, and atomic scale, reaching an understanding of the compositional evolution of cementite and ferrite even at the very lowest level. All the obtained information, including thermal analysis and thermodynamic simulation results, allow to identify the deformation mechanisms present during cold drawing and at the delamination process. In addition, the compositional evolution of cementite and ferrite is also clarified by atom probe tomography.

3D mesoscale modelling of recycled aggregate concrete

Mesoscale modelling is a promising tool to analyse the phenomenological causes of mechanical failure of recycled aggregate concrete (RAC). Most mesoscale models of RAC are two-dimensional (2D) and 3D models are lacking. This prevents the full understanding of the behaviour of each mesoscopic phase. This paper presents a 3D model of RAC and analyses its mechanical behaviour based on mesoscale phenomena. The analyses concern three types of tests carried out in concrete specimens: uniaxial compressive, uniaxial tensile and splitting tensile strength tests. The mechanical response and fracture behaviour of each mesoscopic phase is analysed in relation to the response of RAC specimen. The comparisons between uniaxial behaviour of 3D and 2D modelling as well as between uniaxial and splitting tensile behaviour of RAC are presented. The main findings of the paper are: (1) the full curve of uniaxial stress-strain response from 3D modelling is comparable to that from 2D modelling for RAC (therefore, 2D models are validated for RAC modelling); (2) the overall behaviour of the concrete elements remains elastic regardless of the fracture development in the ITZ and attached mortar of recycled aggregates until damage propagates to the new mortar matrix; (3) a relationship between modelled uniaxial and splitting strength of RAC is established and validated.

Models for Simulation of Fractal-like Particle Clusters with Prescribed Fractal Dimension

This review article delves into the growing recognition of fractal structures in mesoscale phenomena. The article highlights the significance of realistic fractal-like aggregate models and efficient modeling codes for comparing data from diverse experimental findings and computational techniques. Specifically, the article discusses the current state of fractal aggregate modeling, with a focus on particle clusters that possess adjustable fractal dimensions (Df). The study emphasizes the suitability of different models for various Df–intervals, taking into account factors such as particle size, fractal prefactor, the polydispersity of structural units, and interaction potential. Through an analysis of existing models, this review aims to identify key similarities and differences and offer insights into future developments in colloidal science and related fields.

The Impact of Charge-Neutral Atmospheric Propagation Path on the Altimetry Performance of Interferometric Radar Altimeter

Abstract The interferometric radar altimeter (IRA) is an innovative remote sensing sensor that enables the observation of mesoscale and submesoscale (meso–submesoscale) ocean dynamic phenomena. The charge-neutral atmosphere introduces path delay and bending in signal propagation. In this study, three types of sea surface height (SSH) errors caused by charge-neutral atmosphere propagation path for IRA were identified: differential delay error (DDE), path delay error (PDE), and path bending error (PBE). Among them, DDE exhibits a proportionality to the negative zenith neutral delay (ZND) and demonstrates a significant increase with the incident angle; PDE is solely reliant on the ZND; PBE is like DDE in trend and magnitude resembling a ramp. Intriguingly, PBE exhibits insensitivity to variations in the charge-neutral atmosphere, behaving more like a systematic error. Theoretically, PBE leads to an increase in the SSH error of about 1.2 cm at far range for SWOT. The ZND spectrum fitted from the Jason-3 zenith delay correction data is additionally utilized to simulate the spatial distribution of ZND anomaly within the SWOT observation swaths. Then, the impact of PDE anomaly (PDEA), PBE, and DDE anomaly (DDEA) on the observation performance of SWOT is also considered in conjunction with SSH data provided by HYCOM. The findings indicate that both PDEA and PBE significantly reduce IRA’s performance in oceanic phenomena, while the impact of DDEA can be disregarded. The PBE can distort the sea surface trend and increases the mean sea level within the range, requiring further attention. Significance Statement This paper focuses on how the signal path affects the accuracy of measuring sea surface height (SSH) using the interferometric radar altimeter (IRA) in the charge-neutral atmosphere. The present paper defines three types of SSH errors caused by propagation path to IRA, namely, differential delay error (DDE), path delay error (PDE), and path bending error (PBE). Three types of SSH errors will make a significant impact on the altimetry performance of IRA. Among them, PDE spatial anomaly and PBE will reduce the expected observation performance of the meso–submesoscale ocean phenomena in IRA’s swath, while PBE has not received enough attention in previous studies. Therefore, the research will provide a theoretical basis for IRA to correct the above SSH errors.

A lightweight deep learning model for ocean eddy detection

Ocean eddies are typical oceanic mesoscale phenomena that are numerous, widely distributed and have high energy. Traditional eddy detection methods are mainly based on physical mechanisms with high accuracy. However, the large number of steps and complex parameter settings limit their applicability for most users. With the rapid development of deep learning techniques, object detection models have been broadly used in the field of ocean remote sensing. This paper proposes a lightweight eddy detection model, ghost eddy detection YOLO (GED-YOLO), based on sea level anomaly data and the “You Only Look Once” (YOLO) series models. The proposed model used ECA+GhostNet as the backbone network and an atrous spatial pyramid pooling network as the feature enhancement network. The ghost eddy detection path aggregation network was proposed for feature fusion, which reduced the number of model parameters and improved the detection performance. The experimental results showed that GED-YOLO achieved better detection precision and smaller parameter size. Its mAP was 95.11% and the parameter size was 22.56 MB. In addition, the test experiment results showed that GED-YOLO had similar eddy detection performance and faster detection speed compared to the traditional physical method.

Analysis of the Gálvez–Davison Index for the Forecasting Formation and Evolution of Convective Clouds in the Tropics: Western Cuba

The Gálvez–Davison Index (GDI) is an atmospheric stability index recently developed to improve the prediction of thunderstorms and shallower types of moist convection in the tropics. Because of its novelty, its use for tropical regions remains largely unexplored. Cuba is a region that suffers extreme weather events, such as tropical storms and hurricanes, some of them worsened by climate change. This research analyzes the effectiveness of the GDI in detecting the potential for convective cloud development, using forecast data from the Weather Research and Forecasting (WRF) model for Western Cuba. To accomplish this, here, we evaluated the performance of the GDI in ten study cases from the dry and wet seasons. As part of our study, we researched how GDI correlates with brightness temperatures (BTs) measured using GOES-16. In addition, the GDI results with the WRF model are compared with results using the Global Forecast System (GFS). Our results show a high correlation between the GDI and BT, concluding that the GDI is a robust tool for forecasting both synoptic and mesoscale convective phenomena over the region studied. In addition, the GDI is able to adequately forecast stability conditions. Finally, the GDI values computed from the WRF model perform much better than those from the GFS, probably because of the greater horizontal resolution in the WRF model.

Sea-land breezes in the Guangdong- Hong Kong- Macau Greater Bay Area coastal zone from 2013 to 2022

Sea-land breezes (SLBs11SLB, sea-land breeze; GBA, Greater Bay Area) are mesoscale circulation phenomena that significantly affect weather and air quality in coastal regions. The Guangdong–Hong Kong–Macau Greater Bay Area (GBA), one of the most densely populated and economically developed regions in China, experiences frequent SLB intrusion. In this study, we examined the SLB features in the GBA from 2013 to 2022 based on observations, focusing particularly on the influence of synoptic conditions on SLB behaviour and their interplay. The annual average daily sea and land wind speeds ranged from 2.8 to 3.2 m/s and 1.9 to 2.2 m/s, and from 1.8 to 2.2 m/s and 1.6 to 1.9 m/s in the eastern and western GBA, respectively. The start time of sea (land) winds first appeared at 0000 (1100) UTC and finally occurred at 0900 (2000) UTC. Climatologically, SLBs were observed more frequently when southerly synoptic winds prevailed in the GBA. SLBs were stronger under synoptic northerly winds, but weaker when synoptic southerly winds prevailed. The diurnal evolution amplitude of SLB strength was much more significant under the control of synoptic northerly winds. In addition, SLBs modified synoptic winds, leading to the distortion of the observed diurnal wind patterns. A southerly increment in synoptic winds was observed from noon to early night on SLB days, whereas a northerly increment was observed from midnight to morning. Northerly synoptic winds weakened or even transformed into southerly winds in the afternoon, whereas southerly synoptic winds were found to decrease or even turn northerly at night. Our study provides an integrated understanding of SLBs in the GBA, especially regarding the behaviour of SLBs under various synoptic conditions and their modification on synoptic winds that reshape the observed wind patterns. These results provide useful information for weather simulations in coastal areas.

Unravelling the improved permeability of wet granular packed bed via embedding dry particles: From multi-scale mechanisms to combustion enhancements

The permeability of wet granular packed beds inevitably deteriorate as their thickness increases. Therefore, a wide variety of technologies have been developed to improve bed permeability. Return fines embedding (RFE) is a promising technology owing to its simplicity and efficiency. Improved bed permeability using RFE was validated by increasing the moisture content from 7.5% to 9.0%. Return fines larger than 3 mm, particularly those between 4 mm and 5 mm, played a key role in improving bed permeability, and the corresponding mass ratio to the total weight of return fines is not expected to exceed 60%. It was also found that the return fines embedded in the lower zone had better permeability than those embedded in the upper zone owing to the formation of an overly wet zone close to the bottom. As a result, the particle-scale porous structures and the meso-scale moisture transfer phenomena were investigated to unravel the bed-scale improved permeability. In addition to the multi-scale mechanisms, the combustion enhancements in the wet granular packed bed by applying the RFE strategy were visually demonstrated.

The Seasonality of Eddy-Induced Chlorophyll-a Anomalies in the Kuroshio Extension System

The Kuroshio Extension (KE) System exhibits highly energetic mesoscale phenomena, but the impact of mesoscale eddies on marine ecosystems and biogeochemical cycling is not well understood. This study utilizes remote sensing and Argo floats to investigate how eddies modify surface and subsurface chlorophyll-a (Chl-a) concentrations. On average, cyclones (anticyclones) induce positive (negative) surface Chl-a anomalies, particularly in winter. This occurs because cyclones (anticyclones) lift (deepen) isopycnals and nitrate into (out of) the euphotic zone, stimulating (depressing) the growth of phytoplankton. Consequently, cyclones (anticyclones) result in greater (smaller) subsurface Chl-a maximum (SCM), depth-integrated Chl-a, and depth-integrated nitrate. The positive (negative) surface Chl-a anomalies induced by cyclones (anticyclones) are mainly located near (north of) the main axis of the KE. The second and third mode represent monopole Chl-a patterns within eddy centers corresponding to either positive or negative anomalies, depending on the sign of the principal component. Chl-a concentrations in cyclones (anticyclones) above the SCM layer are higher (lower) than the edge values, while those below are lower (higher), regardless of winter variations. The vertical distributions and displacements of Chl-a and SCM depth are associated with eddy pumping. In terms of frequency, negative (positive) Chl-a anomalies account for approximately 26% (18%) of the total cyclones (anticyclones) across all four seasons. The opposite phase suggests that nutrient supply resulting from stratification differences under convective mixing may contribute to negative (positive) Chl-a anomalies in cyclone (anticyclone) cores. Additionally, the opposite phase can also be attributed to eddy stirring, trapping high and low Chl-a, and/or eddy Ekman pumping. Based on OFES outputs, the seasonal variation of nitrate from winter to summer primarily depends on the effect of vertical mixing, indicating that convective mixing processes contribute to an increase (decrease) in nutrients during winter (summer) over the KE.

Mesoscale Phenomena & their Contribution to the Global Response: A Focus on the Magnetotail Transition Region & Magnetosphere-Ionosphere Coupling

Mesoscale Phenomena & their Contribution to the Global Response: A Focus on the Magnetotail Transition Region & Magnetosphere-Ionosphere Coupling

Nested computational fluid dynamic modeling of mean turbulent quantities estimation in complex topography using AROME-SIMRA

This paper presents a validation of the AROME-SIMRA model, which is a nested computational fluid dynamics model that simulates both mesoscale and microscale phenomena. To validate the model, we analyzed 47 h of mean flow data collected by 13 three-dimensional sonic anemometers. These anemometers were mounted on tall masts located near the shoreline of Sulafjord, with heights ranging from 12m to 95m above the ground. Due to the difficulty measuring wind along the bridge span, analyzing flow conditions for the construction of a bridge that spans a vast fjord is a difficult process. Therefore, the primary objective of this study is to validate a nested macroscale–microscale model. This model will be utilized to analyze flow conditions across the span of the proposed bridge crossing in Sulafjord. The study explores the deviation between the measured and the simulated mean turbulence flow characteristics. Only records with the mean wind of 12ms−1 and above at SulaNW met-mast are considered due to their relevance in buffeting response, which led to the identification of a limited number of sectors representative of strong wind conditions. Mean wind speed comparisons show a minimum correlation of 0.6 and a maximum of 0.9 for all the anemometers analyzed. For wind directions, a low correlation between observation and numerical simulation is obtained at SulaSW met-mast located southwest of Sulafjord. A high Angle of Attack is obtained for both simulation and measurements. However, the correlation is dependent on the mast location, wind direction, and anemometer height. Along the bridge span, the flow is largely horizontal for the northwestern flow.

Reinforcement learning based hybrid bond-order coarse-grained interatomic potentials for exploring mesoscale aggregation in liquid-liquid mixtures.

Exploring mesoscopic physical phenomena has always been a challenge for brute-force all-atom molecular dynamics simulations. Although recent advances in computing hardware have improved the accessible length scales, reaching mesoscopic timescales is still a significant bottleneck. Coarse-graining of all-atom models allows robust investigation of mesoscale physics with a reduced spatial and temporal resolution but preserves desired structural features of molecules, unlike continuum-based methods. Here, we present a hybrid bond-order coarse-grained forcefield (HyCG) for modeling mesoscale aggregation phenomena in liquid-liquid mixtures. The intuitive hybrid functional form of the potential offers interpretability to our model, unlike many machine learning based interatomic potentials. We parameterize the potential with the continuous action Monte Carlo Tree Search (cMCTS) algorithm, a reinforcement learning (RL) based global optimizing scheme, using training data from all-atom simulations. The resulting RL-HyCG correctly describes mesoscale critical fluctuations in binary liquid-liquid extraction systems. cMCTS, the RL algorithm, accurately captures the mean behavior of various geometrical properties of the molecule of interest, which were excluded from the training set. The developed potential model along with the RL-based training workflow could be applied to explore a variety of other mesoscale physical phenomena that are typically inaccessible to all-atom molecular dynamics simulations.