JGR–Atmospheres

Satellite remote sensing serves as a crucial means to acquire cloud physical parameters. However, existing official cloud products from the advanced geostationary radiation imager (AGRI) onboard the Fengyun-4A geostationary satellite lack spatiotemporal continuity and important micro-physical properties. In this study, an image-based transfer learning ResUnet (TL-ResUnet) model was applied to realize all-day and high-precision retrieval of cloud physical parameters from AGRI thermal infrared measurements. Combining the observation advantages of geostationary and polar-orbiting satellites, the TL-ResUnet model was pre-trained with official cloud products from advanced Himawari imager (AHI) and transfer-trained with official cloud products from moderate resolution imaging spectroradiometer (MODIS), respectively. For comparison, a pixel-based transfer learning random forest (TL-RF) model was trained using the equally distributed data sets. Taking MODIS official products as the benchmarks, the TL-ResUnet model achieved an overall accuracy of 79.82% for identifying cloud phase and root mean squared errors of 1.99 km, 7.11 μm, and 12.87 for estimating cloud top height, cloud effective radius, and cloud optical thickness, outperforming the precision of AGRI and AHI official products. Compared to the TL-RF model, the TL-ResUnet model utilized the spatial information of clouds to significantly improve the retrieval performance and achieve more than a 6-fold increase in speed for single full-disk retrieval. Moreover, AGRI TL-ResUnet products with spatiotemporal continuity and high precision were used to accurately describe the spatial distribution characteristics of cloud fractions and cloud properties over the Tibetan Plateau, and provide the diurnal variation of cloud cover and cloud properties across different seasons for the first time.

Arctic permafrost thaw holds the potential to drastically alter the Earth's surface in Northern high latitudes. We utilize high-resolution large eddy simulations to investigate the impact of the changing surfaces onto the neutrally stratified atmospheric boundary layer (ABL). A stochastic surface model based on Gaussian Random Fields modeling typical permafrost landscapes is established in terms of two land cover classes: grass land and open water bodies, which exhibit different surface roughness length and surface sensible heat flux. A set of experiments is conducted where two parameters, the lake areal fraction and the surface correlation length, are varied to study the sensitivity of the boundary layer with respect to surface heterogeneity. Our key findings from the simulations are the following: The lake areal fraction has a substantial impact on the aggregated sensible heat flux at the blending height where surface heterogeneities become horizontally homogenized. The larger the lake areal fraction, the smaller the sensible heat flux. This result gives rise to a potential feedback mechanism. When the Arctic dries due to climate heating, the interaction with the ABL may accelerate permafrost thaw. Furthermore, the blending height shows significant dependency on the correlation length of the surface features. A longer surface correlation length causes an increased blending height. This finding is of relevance for land surface models concerned with Arctic permafrost as they usually do not consider a heterogeneity metric comparable to the surface correlation length.

Biogenic volatile organic compounds (BVOCs) are regarded as important precursors for ozone and secondary organic aerosol, mainly from vegetation emissions. In the context of the expanding trend of vegetation greening, the development of high-precision vegetation data and accurate BVOC emission estimates are essential to develop effective air pollution control measures. In this study, by integrating the multi-source vegetation cover data, we established a high-resolution vegetation distribution (HRVD) data set to develop a high spatio-temporal resolution emission inventory and investigated the impact of different land cover data sets on emission simulation and impact of land cover change on BVOC emissions during 2001–2020. The annual total BVOC emissions in China for 2020 was 15.66 Tg, which were mainly from trees. The emissions simulated by CNLUCC and MODIS data sets were 1.53% and 1.72% higher than those simulated by HRVD data sets, respectively. The spatial distribution of emission differences was consistent with that of land cover differences. The simulated BVOC emissions by the HRVD data set had the best accuracy as they improved the bias between modeling and observation from 69.06% to 65.35% and decreased the underprediction of observations by a factor of 2.13 compared with simulation by MEGAN default vegetation data. The annual BVOC emissions caused by changing vegetation distribution and LAIv (LAI of vegetation covered surfaces) enhanced at a rate of 72.06 Gg yr−1 during 2001–2020. LAIv was the main driver of emission variations. The total OH reactivity of the resulted BVOC emissions increased at a rate of 1.59 s−1 yr−1, with isoprene contributed the most.

The return stroke of cloud-to-ground (CG) lightning is an impulsive radiator of very low-frequency/low-frequency (VLF/LF) electromagnetic signals allowing for the remote sensing of lower ionosphere over large spatial coverage. In this study, we examined the LF magnetic fields measured in Malacca, Malaysia, to probe reflection heights of the lower ionosphere near the equator on three different nights in 2021. The results show that the virtual ionospheric height at nighttime typically ranged from 82.0 to 90.0 km, with a mean value of 85.3 km. Our measurements also revealed significant variations in the virtual ionospheric height across different regions over a spatial scale of about 800 km. The maximum height difference was about 5.0 km. Moreover, the fluctuation characteristics are observed in both estimated ionospheric height and calculated peak reflection ratio during similar periods. This fluctuation may be related to atmospheric gravity waves in the nighttime ionosphere. In addition, we compared the virtual ionospheric height estimated from CG strokes of different polarities, and the results showed that the virtual reflection height for positive CG strokes is lower than that for negative ones.

A persistent shortwave radiative bias of Southern Ocean (SO) clouds in climate models is strongly associated with incorrect cloud phase representation, which impacts precipitation. Measurements characterizing precipitation in low-level mixed-phase clouds, which frequently form over the SO, are rare, and our understanding of precipitation efficacy within these clouds remains limited. The simulated surface precipitation bias has an indirect effect on determining global climate sensitivity and a direct impact on the hydrological cycle. This study investigates the representation of low clouds, cloud variability, and precipitation statistics over the SO in real-case Icosahedral Nonhydrostatic (ICON) simulations at the kilometer scale. The simulations are contrasted with 48 hr of continuous shipborne observations of open and closed-cell stratocumuli, south of Tasmania. Our simulations show the significance of heavily rimed particle formation, their in-cloud growth, and subcloud melting to capture the observed cloud-precipitation vertical structure. In addition, supercooled drizzle formation impacts the vertical structure and precipitation statistics. ICON captures the observed intermittency of precipitation even at a standard vertical resolution of 200 m in the boundary layer but only captures the observed sparse distribution of intense precipitation (>1 mm hr−1) when the maximum vertical resolution is reduced to 100 m. However, the simulations of the 2-day accumulated precipitation and the radiative effect are largely insensitive to the vertical resolution. The cloud reflectivity of the broken cloud deck is underestimated due to negative biases in cloud optical depth.

Air pollution caused by various anthropogenic activities and biomass burning continues to be a major problem in India. To assess the effectiveness of current air pollution mitigation measures, we used a 3D global chemical transport model to analyze the projected optical depth of carbonaceous aerosol (AOD) in India under representative concentration pathways (RCP) 4.5 and 8.5 over the period 2000–2100. Our results show a decrease in future emissions, leading to a decrease in modeled AOD under both RCPs after 2030. The RCP4.5 scenario shows a 48%–65% decrease in AOD by the end of the century, with the Indo-Gangetic Plain (IGP) experiencing a maximum change of ∼ ${\sim} $25% by 2030 compared to 2010. Conversely, RCP8.5 showed an increase in AOD of ∼ ${\sim} $29% by 2050 and did not indicate a significant decrease by the end of the century. Our study also highlights that it is likely to take three decades for current policies to be effective for regions heavily polluted by exposure to carbonaceous aerosols, such as the IGP and eastern India. We emphasize the importance of assessing the effectiveness of current policies and highlight the need for continued efforts to address the problem of air pollution from carbonaceous aerosols, both from anthropogenic sources and biomass burning, in India.

Elevated surface ozone levels are often detected in the New York metropolitan area during summertime. Moreover, surface ozone in this region exhibits sharp spatial gradients and distinctive diurnal cycles under the influence of complex boundary layer circulations induced by the intricate coastal geometry. This study examines how surface ozone is impacted by local circulations spatially and temporally under different temperature scenarios (all summer days, hot summer days, and extreme heat days) with the help of cluster-based meteorological conditions during the summertime of 2017–2019. The most polluted days are found to be highly associated with hot sea breeze days with weak background flow. When sea breeze development in the New York Bight is delayed and its penetration north is intercepted by the dominant westerlies during hot summer days, daily maximum 8-hr average ozone (DMA8) in some ozone hot spots of New York City (NYC) and the south shore of Connecticut (CT) typically drops 9–10 ppb under comparable temperature levels. The average regional decrease of DMA8 for NYC and coastal CT is 6.7 and 8.3 ppb, respectively. Furthermore, we conclude that a change in early morning meridional wind direction is the most critical meteorological characteristic in controlling sea breeze onset type and helping modulate ozone exceedances in the region during extreme hot days when ozone exceedances are expected to be very common. The conclusion is further demonstrated with two case studies during the Long Island Sound Tropospheric Ozone Study 2018 field campaign.

This study derives polarimetric radar vertical profiles and microphysical retrievals for 25 Synoptic Snow (SS) and 23 Lake Effect Snow (LES) cases using the Range-Defined Quasi-Vertical Profiles (RD-QVP), Columnar Vertical Profiles (CVP), and Process-oriented Vertical Profiles (POVP) methods. For all vertical profile techniques, SS cases exhibit a near-linear increase in reflectivity from −30 to 0°C whereas Z DR and K dp locally peak in the dendritic growth layer. LES cases universally exhibit negative Z DR , rather high Z, negligible K dp , and near-unity ρ hv . Ground measurements from the past OWLeS campaign provide direct evidence that conical graupel may strongly affect these polarimetric measurements in LES bands. Aggregation efficiencies for SS cases are estimated by optimizing the theoretical number concentration (N t ) and mean volume diameter (D m ) steady-state vertical profiles against radar-retrieved profiles derived from 20 of the 25 synoptic storm RD-QVPs. The median estimated aggregation efficiency is approximately 0.15 with a relatively narrow interquartile range that spans from 0.1 to just over 0.2. Values of optimized aggregation efficiencies are nearly independent of the assumed gamma distribution shape parameter. These results are used to derive temperature-dependent, climatological steady-state relations for vertical profiles of N t , D m , and liquid-equivalent snowfall rates. These results can be used in numerical weather prediction model aggregation parameterizations and can also provide climatologically representative vertical profiles of radar and microphysical quantities.

Surface fluxes and states can recur and remain consistent across various spatial and temporal scales, forming space-time patterns. Quantifying and understanding the observed patterns is desirable, as they provide information about the dynamics of the processes involved. This study introduces the empirical spatio-temporal covariance function and a corresponding parametric covariance function as tools to identify and characterize spatio-temporal patterns in remotely sensed fields. The method is demonstrated using 2 km hourly GOES-16/17 land surface temperature (LST) data over the Contiguous United States by splitting the area into 1.0° × 1.0° domains. The summer day-time LST ESTCFs for 2018 to 2022 are derived for each domain, and a parametric covariance model is fitted. Clustering analysis is applied to detect areas with similar spatio-temporal LST patterns. Six main zones within CONUS are identified and characterized based on their variance and temporal and spatial characteristic length scales (i.e., scales for which the temperature variations are temporally and spatially related), respectively: (a) Eastern plains with 3 K2, ∼6 hr, and 0.15°, (b) Gulf of California with 60 K2, ∼8 hr, and 0.34°, (c) mountains and coasts transition 1 with 16 K2, ∼11 hr, and 0.25°, (d) central US, Midwest, and South cities with 5.5 K2, ∼8 hr, and ∼0.2°, (e) mountains and coasts transition 2 with ∼10 K2, ∼8 hr, and 0.2°, and (f) largest mountains and coastlines with ∼19 K2, ∼13 hr, and 0.3°. The tools introduced provide a pathway to formally identify and summarize the spatio-temporal patterns observed in remotely sensed fields and relate those to more complex processes within the Soil-Vegetation-Atmosphere System.

On 18 December 2022, Hawaiian Airlines flight HA35 encountered severe turbulence without warning in a cloud-free height. We reproduced this incident using the Weather Research and Forecasting Model (WRF) at a convection-permitting resolution. We found that this case of tropical upper-level turbulence occurred primarily due to the fast-growing convective tower in the unstable layer created by gravity wave breaking. At lower altitudes, a mesoscale convective system (MCS) caused a decrease in wind speed in both upstream and downstream regions. At upper levels, a large-scale jet descended and accelerated after flowing over the top of the MCS, which acted like a barrier and produced a situation similar to a downslope windstorm due to mountain terrain. Upper-level turbulence is 2–3 km higher than the top of the MCS. The critical level above the jet and the locally self-induced critical level created the locally enhanced descending jet stream, which destabilized the flow through Kelvin–Helmholtz instabilities. The convective tower existed near the flight route and played an important role in triggering turbulence in the unstable environment through its convective gravity waves.

Recently, more advanced synchronous global-scale satellite observations, the Soil Moisture Active Passive enhanced Level 3 (SMAP L3) soil moisture product and the Orbiting Carbon Observatory 2 (OCO-2) solar-induced chlorophyll fluorescence (SIF) product, provide an opportunity to improve the predictive understanding of both water and carbon cycles in land surface modeling. The Simplified Simple Biosphere Model version 4 (SSiB4) was coupled with the Top-down Representation of Interactive Foliage and Flora Including Dynamics Model (TRIFFID) and a mechanistic representation of SIF. Incorporating dynamic vegetation processes reduced global SIF root-mean-squared error (RMSE) by 12%. Offline experiments were conducted to understand the water and carbon cycles and their interactions using satellite data as constraints. Results indicate that soil hydraulic properties, the soil hydraulic conductivity at saturation (Ks) and the water retention curve, significantly impact soil moisture and SIF simulation, especially in the semi-arid regions. The wilting point and maximum Rubisco carboxylation rate (Vmax) affect photosynthesis and transpiration, then soil moisture. However, without atmospheric feedback processes, their effects on soil moisture are undermined due to the compensation between soil evaporation and transpiration. With optimized parameters based on SMAP L3 and OCO-2 data, the global RMSE of soil moisture and SIF simulations decreased by 15% and 12%, respectively. These findings highlight the importance of integrating advanced satellite data and dynamic vegetation processes to improve land surface models, enhancing understanding of terrestrial water and carbon cycles.

Meteorological observations of surface air temperature have provided fundamental data for climate change detection and attribution. However, the weather stations are unevenly distributed, and are still very sparse in remote regions. The possible sampling error is well known, but not well quantified because we are lack of the adequate and regularly distributed measurements. The high resolution of satellite land surface temperature retrieval during night time provide a nice proxy for near surface temperature as both temperatures controlled by surface longwave radiative cooling and the nocturnal temperature inversion depress land-atmosphere turbulent exchange. The sampling error of mean value and trend were assessed by comparing station point measurements (pixel of ∼0.01°) with grid (1°) mean and national mean from 2001 to 2021. This method permits us to make the first assessment of under-sampling error and spatial representative error on both national mean and trend of air temperature during nighttime collected at ∼2,400 weather stations over China. The sampling error in national mean temperature is more than 3°C. The under-sampling error due to lack of observation explains two thirds and the spatial representative error due to the difference between station and grid/regional mean elevation contribute the other one third. The sampling error in trend account for one third of the national mean trend. The urban heat island effect associated with urbanization around the weather stations (spatial representative error) can explain four fifths of the sampling error in trend, which is consistent with existing studies based on air temperature collected at paired weather station.

To investigate the combined impacts of anthropogenic and biogenic emissions on the formation of secondary organic aerosols (SOA), SOA molecular tracers, their corresponding volatile organic compound precursors, and other air pollutants were measured online during the winter and summer seasons of 2022 in an industrial city, Zibo, China. The results indicate that the average concentrations of SOA tracers were 16.1 ± 9.8 ng m−3 in winter and 99.4 ± 57.2 ng m−3 in summer. During winter, anthropogenic SOA (ASOA, the sum of SOA derived from naphthalene and mono-aromatic volatile organic compounds) dominated, whereas isoprene SOA (SOAI) prevailed in summer. Correlation analysis between SO4 2− and both SOAI and high-order monoterpene SOA tracers (SOAM-H) (R = 0.46–0.72, p < 0.001) revealed that higher aerosol acidity facilitated the formation of SOAI and SOAM-H, with SO2 emissions playing a significant role in leading to higher acidity. Most biogenic SOA (BSOA) tracers exhibited a significant positive correlation with NO3 −, particularly in winter, implying the remarkable influence of NO x emissions on BSOA formation. The levels of BSOA tracers increased with NH3, indicating that NH3 can enhance the formation of BSOA. In summer, SOA formation correlated with O x (O x  = O3 + NO2), indicating the substantial impact of atmospheric oxidizing capacity on SOA formation. During winter, aerosol liquid water content (ALWC) correlated well with SOAI tracers (i.e., 3-hydroxyglutaric acid (3-HGA) and 3-hydroxy-4,4-dimethylglutaric acid (3-HDMGA)), and 2,3-dihydroxy-4-oxopentanoic acid (DHOPA) (R > 0.5, p < 0.001), indicating the important contribution of aqueous-phase formation of SOA. These findings underscore the significant role of anthropogenic pollutant emissions in the formation of ASOA and BSOA in urban environments.

This study uses a unique combination of geostationary and low-Earth orbiting satellite-based lightning and precipitation observations, respectively, to examine the evolution of deep convection during the tropical cyclone (TC) lifecycle. The study spans the 2018–2021 Atlantic Basin hurricane seasons and is unique as it provides the first known analysis of total lightning (intra-cloud and cloud-to-ground) observed in TCs through their extratropical transition and post-tropical cyclone (PTC) phases. We consider the TC lifecycle stage, geographic location (e.g., land, coast, and ocean), shear strength, and quadrant relative to the storm motion and environmental shear vectors. Total lightning maxima are found in the forward right quadrant relative to storm motion and downshear of the TC center, consistent with previous studies using mainly cloud-to-ground lightning. Increasing environmental shear focuses the lightning maxima to the downshear right quadrant with respect to the shear vector in tropical storm phases. Vertical profiles of radar reflectivity from the Global Precipitation Measurement mission show that super-electrically active convective precipitation features (>75 flashes) within the PTC phase of TCs have deeper mixed phase depths and higher reflectivity at −10°C than other phases, indicating the presence of more intense convection. Differences in the net convective behavior observed throughout TC evolution manifest in both the TC-scale frequency of lightning-producing cells and the intensity variations amongst individual convective cells. The combination of continuous lightning observations and precipitation snapshots improves our understanding of convective-scale processes in TCs, especially in PTC phases, as they traverse the tropics and mid-latitudes.

Secondary organic aerosol (SOA) formed through the atmospheric transformation of organic vapors constitutes a significant portion of fine particulate matter or PM2.5. While recent laboratory studies underscore the importance of intermediate-volatility organic compounds (IVOCs) as key precursors to SOA, field observations that recognize the role of both volatile organic compounds (VOCs) and IVOCs in SOA formation remain scarce. In this study, we conducted concurrent measurements of VOCs and IVOCs in ambient air at urban and suburban sites in Guangzhou during a PM2.5 pollution event in winter 2021. The results reveal that between 12:00–15:00 local time, the photochemically adjusted initial concentrations of VOCs at both sites were approximately 7 times higher than that of IVOCs. However, the SOA formation potential (SOAFP) of primary hydrocarbon IVOCs exceeded that of VOCs by over 3–4 times. Receptor modeling results further indicated that while ship emissions contributed to less than 10% of the C2–C22 primary hydrocarbons concentration (VOCs + primary carbonaceous IVOCs), they accounted for the most significant source (approximately 40%) of SOA formation. This study highlights the substantial role of IVOCs in SOA formation and emphasizes the importance of future PM2.5 pollution control measures targeting major IVOCs contributors, such as ship emissions in harbor cities.

No abstract is available for this article.

The parameterizations of air-sea turbulent heat flux are one of the major bottlenecks in atmosphere-ocean coupled model development, which play a crucial role in sea surface temperature (SST) prediction. Recently, neural networks start to be applied for the development of parameterizations of interface turbulent heat flux. However, these new parameterizations are primairily developed for specific regions and have not been tested in real atmosphere-ocean coupled models. In this study, we propose a new air-sea heat flux parameterization using a physical-informed neural network (PINN) based on multiple observational data sets worldwide. Evaluated with an independent observation data set, it is shown that the PINN can significantly reduce the RMSE of latent heat flux by at least about 48.6% compared to three traditional bulk formulas. Moreover, the PINN can be flexibly updated with new observational data by transfer learning. To test the performance of the new parameterization in realistic application, we implement the PINN into a global ocean-atmosphere coupled model and make seasonal forecasts for the first time. The PINN markedly reduce the errors of equatorial SST forecast, indicating a good performance of the PINN-based air-sea turbulent heat flux scheme. Noticeably, due to limited observational data, the NN-based parameterizations tend to underestimate heat flux at high wind speeds compared with bulk formula-based parameterizations. With more data available at extreme conditions, the PINN can be improved via transfer learning and need to be futher evaluated. This study suggests that PINN-based air-sea heat flux parameterization is promising to improve SST simulation.

Arctic cyclones are key drivers of sea ice and ocean variability. During the 2019–2020 Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition, joint observations of the coupled air-ice-ocean system were collected at multiple spatial scales. Here, we present observations of a strong mid-winter cyclone that impacted the MOSAiC site as it drifted in the central Arctic pack ice. The sea ice dynamical response showed spatial structure at the scale of the evolving and translating cyclonic wind field. Internal ice stress and ocean stress play significant roles, resulting in timing offsets between the atmospheric forcing and the ice response and post-cyclone inertial ringing in the ice and ocean. Ice motion in response to the wind field then forces the upper ocean currents through frictional drag. The strongest impacts to the sea ice and ocean from the passing cyclone occur as a result of the surface impacts of a strong atmospheric low-level jet (LLJ) behind the trailing cold front and changing wind directions between the warm-sector LLJ and post cold-frontal LLJ. Impacts of the cyclone are prolonged through the coupled ice-ocean inertial response. Local impacts of the approximately 120 km wide LLJ occur over a 12 hr period or less and at scales of a kilometer to a few tens of kilometers, meaning that these impacts occur at combined smaller spatial scales and faster time scales than most satellite observations and coupled Earth system models can resolve.

Dry- and wet-bulb temperature (T d and T w ) are usually to define heatwaves (HWs) which have been enhanced under anthropogenic climate change (ACC) and urbanization. However, responses of various types of HWs (i.e., dry HWs, only high T d ; humid HWs, only high T w ; hybrid HWs, both high T d and T w ; total HWs, high T d or T w ), to ACC and urbanization remain unknown. In this study, both observations and simulations show significantly increasing occurrence probability of total HWs over China during 1971–2020, whereas this increase is mainly reflected in hybrid HWs, followed by dry HWs and humid HWs. 68.2%–93.0% of the observed increases in the above four types of HWs can be attributed to ACC; on the other hand, urbanization tends to suppress humid HWs but enhance dry HWs, as a result of contributing to the increase of hybrid HWs by 10.9%. Under future ACC, total HWs are projected to be more frequent as expected, which is mainly sourced from the increasing hybrid HWs because dry/humid HWs are projected to be steady/downward. As a consequence, urban population exposure to ACC-induced total HWs would remarkably increase to 83.55 billion person-days by the 2090s, 89.5% of which can be attributed to hybrid HWs. Urbanization would amplify this population exposure of ACC-induced hybrid HWs from 74.79 billion person-days to 110.9 billion person-days. Our results underscore the importance of improving understanding of hybrid HWs in urban areas and developing targeted adaptation planning on a warmer planet.

As a significant macrophysical property, cloud horizontal scales play a role in cloud radiation, precipitation and vertical cloud overlap. Until now, however, the mechanisms behind the variations in cloud scale distribution have received far less attention. This study utilizes active satellite data from 2007 to 2016 to investigate the spatiotemporal distribution of cloud horizontal scales, and explains the variations through two meteorological factors: wind shear and atmospheric stability. Cloud scales exhibit a distinct power-law behavior when scale break is not considered, and the power-law exponent β is a characteristic measure of cloud scale distribution. A smaller power-law exponent β indicates a higher frequency of large clouds. During boreal summer season, the amount of large clouds is extremely large south of the 40°S but rather small between 10°S and 20°S. As wind shear decreases or atmospheric stability increases, more large clouds occur globally. The underlying mechanisms might be associated with cloud entrainment which can be promoted by wind shear but inhibited by atmospheric stability. However, our analysis of the impacts of these two factors on cloud scale distribution across different regions and heights reveals that both wind shear and atmospheric stability play dual roles on the values of the exponent β. The potential physical mechanisms, including the effects of precipitation, are further discussed. It is observed that precipitation also exerts a dual impact on the values of the exponent β. These findings underscore the significance of considering the impacts of meteorological factors on cloud scale distribution in numerical weather prediction models.