This research study shows the ability of cloud resolving models (CRM) to simulate Mesoscale Convective Systems (MCS) over the far Eastern Pacific region, off the coast of Colombia and Panama. The simulation period coincides with the newly developed OTREC field campaign (August-September, 2019), which provided enhanced upper-air soundings, NSF/NCAR G-V dropsonde and HIAPER Cloud Radar data to help evaluate the model and diagnose the environmental conditions favoring MCS development. We tested the model sensitivity to three different microphysics schemes: two popular bulk schemes (Thomson and Morrison) and one spectral bin (SBM) scheme. The models are diagnosed on their ability to simulate the observed large-scale and mesoscale environments associated with MCSs development, including the ChocoJet and Caribbean low-level jets, the semi-permanent Panama low, vertical shear, and mid-level diurnal gravity waves. We also examined the vertical distribution of hydrometeors concentrations and diabatic heat and cooling profiles. Results show that not only the SBM represents better the spatial and vertical distribution of precipitation, but also simulates better MCSs characteristics (intensity, duration, organization) and their predominant westward movement. We hypothesize that the success of the SBM in producing better organized and more long-lasting MCS stands in the stronger diabatic heating, related to a top-heavier mass profile that helps support upper-level convergence, and more intense low-level diabatic cooling that helps support stronger gravity currents. OTREC observations and CRM results shed light on the role of MCSs in the generation of enhanced mid-level mesoscale vorticity, which has been related to generation of easterly waves or enhancement of existing ones. Although the SBM is unpractical due to its computation cost (fast version takes about 10-12 times longer), it represents an important step forward in cloud modeling, with suggestive results indicating that SBM improves confidence of the physical basis of the elusive and challenging simulation of realistic tropical MCSs.

Satellite-derived data provide useful information about the rationale of Earth’s functioning. While satellite remote sensing has been regarded as the almost only means for observing the entire Earth in near-real-time, errors in satellite observations have limited their direct usage in applications. Merging two or more data sources has been regarded as a simple but effective way to decrease such errors (e. g. minimizing mean square errors between the observation and truth). The principle of data merging is to combine independent information of each data source, improving over each individual product by canceling out random errors, with effectiveness by the degree of independence over the data sources. In the case of linearly combining data, qualitative assessments of the error (i.e. error variance/covariance and data-truth correlation) are essential to calculate the optimal weight for each candidate product. However, such reference “truth” is rarely available in practical. To overcome this limitation, a triple collocation (TC) technique is often used to estimate data error by using a data triplet without the truth. Despite the usefulness and simplicity of the TC-based error estimation, the inherent assumptions (e.g. error independence) in the approach tend to induce sub-optimal results in the error estimation and/or data combination. There have been also further efforts to address the limitation such as quadruple collocation (QC) using a data quadruple to partially estimate error cross-correlation and single/double instrumental variable methods to lessen the difficulty in obtaining multiple datasets. In this presentation, we review the status of error estimation and data merging approaches based on the collocation methods and then present challenges to be addressed through future research.

Coasts in the Russian Arctic are extremely vulnerable to the ongoing environmental changes. Temporal evolution of their retreat rates is driven by hydrometeorological processes. Permafrost of the coastal bluffs rapidly thaws under the influence of air and water temperature increase. Along with that, sea ice decline results in wave energy increase because of longer wave fetch and ice-free period when the waves are able to erode the unprotected coasts. The combined thermal and wave action is defined as hydrometeorological forcing of coastal erosion. We estimated temporal variability of the air thawing index (sum of annual positive temperatures) reflecting the thermal factor, and the wind-wave energy flux since the 1970s for five sites in both the western and eastern Russian Arctic where observations of coastal erosion rates derived from both field measurements and remotely sensed data are available: Varandey (Pechora Sea), Yamal and Ural coasts of the Baydaratskaya Bay (Kara Sea), Cape Chukochiy (East Siberian Sea) and Lorino (Bering Sea). We further calculated the total hydrometeorological forcing of coastal erosion for the periods with known retreat rates and compared the temporal variability of the two parameters. Comparison of the hydrometeorological forcing shows a link between erosion rates and the hydrometeorological forcing in all the areas. The best correlation is noted for sites where remotely sensed data for relatively long periods were analyzed. For areas with more frequent direct field observations, the variability of the two parameters shows more differences. Such findings imply that while long-term erosion rates are determined by general trends of climate and sea ice extent change in the Arctic seas, coastal retreat in one single year can be driven by local factors, such as lake drainage, random failure of large blocks or peat lenses, exposure and burial of ice bodies, and other reasons. Therefore estimation of mechanisms and trends of coastal erosion in the Russian Arctic should be made based on average retreat rates over relatively long timescales (several years or even decades). Studies on the variability of climate parameters were funded by the RFBR grant 18-05-60300 (S.Ogorodov, N. Shabanova). Studies on rates of coastal erosion were funded by the RSF grant 16-17-00034 (A.Baranskaya, A. Novikova). B.Jones was supported by US National Science Foundation award OISE- 1927553.

The 1,000-fold differences between countries in bringing the 2020 pandemic under control show the consequences of failures to stockpile, a century-old anti-science, anti-expertise streak in American culture, and of tolerating incompetent leadership. The extreme weather in the last twenty years has featured surges in severe windstorms (now 7x as common as before 2008), inland floods, mega heatwaves, hot-dry-windy “fire weather,” and stalled hurricanes. They threaten a human population crash via famine and genocides. Here I adapt the physician’s mental check list to show what a “climate doctor” would consider. First, sensible precautions: beefing up infra­structure to resist floods, burying power lines and creating city-wide battery backup, and relocating people out of flood plains and coastlines. Stockpile. We need economic modeling for emergency measures, what we should have done before the pandemic’s recession. Protect the patient from the usual causes of terminal downhill slides. This is commonly called “stabilizing the patient.” The Civilian Conservation Corps in the 1930s likely prevented civil disorder. Recognize what’s wrong. For climate, the working diagnosis since 1965 is a global-scale overheating caused by the accumulation of CO2, contributed by the annual emissions of fossil fuels. Attacking the root cause may not be fast enough. To evaluate urgency, guess where things are likely headed (prognosis). Climate models are good for estimating slow climate change, but they are only beginning to address the dynamic aspects that can create climate flips and extreme weather. Rule out other problems. Repeatedly search for secondary climate problems, analogous to shock and internal bleeding, that could provoke a fast track to disaster. Formulate a plan of action and explain it to get consent. The annual bump-up in CO2 from emissions is now 50% greater than before 2000. We must now focus on a quick cleanup of the existing CO2 accumulation, as in using a kidney dialysis machine to quickly clean up an aspirin overdose. Gradual overheating is no longer the correct focus for under­standing the risk we now face. We are in the territory where fast tracks to disaster must be forestalled. Something, perhaps the passive drama mindset for the surreal, is preventing effective climate action, even by the knowledge­able.

The discipline of informatics, generically cast as the science and engineering of information system within a socio-technical framework, originating in the middle of last century has undergone generational adaptations as computer hardware, networks and software have evolved. Within the “eScience” era of the last two decades, discipline-specific fields of informatics have flourished, such as geoinformatics, mineral informatics and many more. In fact even in geosciences, there may be few fields of study that have not added an informatics sub-field. Over the same time, efforts at systematizing the common (or core, i.e. discipline neutral) aspects of informatics have been successful: use cases, human-centered design, iterative approaches, information models and more are some of the key elements. However new pressures are being placed on functional and non-functional requirements of information systems: with the now somewhat routine underlying data that are high dimensional, heterogeneous, sparse and with uncertain quality. However, demands have arisen from renewed attention to hub/ server/ cloud-based provision of the application of machine learning, neural networks and artificial intelligence in general. Those methods as implemented in software libraries producing results to be assessed and interpreted (often leading to decisions made) by “humans-in-the-loop”. Informatics, revisited is a possible answer. This presentation features some history of informatics, recent disruptions that need to be addressed, and offers ideas for new directions, with the goal of advancing geoscience in general.

In Peru, a group of scientists are investigating the main Amazonian rivers to understand all those physical variables that define their behavior and ecological importance throughout the Amazon basin. But what happens with this information? How does it connect with the citizens and become mechanisms of valorization and conservation? “Science for Rivers” is a comprehensive educational initiative that seeks to channel all the scientific knowledge generated regarding rivers towards different actors who are linked to them. The communication strategies implemented have translated complex information in a way that is accessible and useful to all audiences. Therefore, many specialized courses and workshops have been developed for different government institutions in order to incorporate this type of knowledge into their development plans. Also, two free access web pages have been developed that show in an interactive way the multiple technical and social variables when intervening in rivers. Likewise, an educational graphic proposal was implemented through social media where information about the rivers is shared with a simple and playful language. Appropriating this knowledge at different levels is fundamental to guarantee policies coherent with this ecosystem, as well as to forge a citizenry that knows, values and cares for its rivers.

Terrestrial wetlands are a highly significant carbon reservoir in North America. Forested wetlands, or swamps, are an important category of North American wetland and include boreal forested peatlands, swamps dominated by needle-leaved trees including Thuja (cedar), Picea (Spruce), Larix (Tamarack) or Taxodium (bald cypress), swamps dominated by broad-leaved trees or shrubs including Fraxinus (Ash), Ulmus (Elm), or Acer (Maple), as well as mangroves. The Second State of the Carbon Cycle Report estimates that forested wetlands may make up ~55% of the total terrestrial wetland area for North America, although estimates vary considerably due to different mapping conventions and classification systems across national and provincial borders, and also due to the ongoing impacts of land use change. Additionally, that report suggests that forested wetlands contain larger total carbon pools than non-forested wetlands, and that forested wetlands effect 53% of the estimated 123 Tg total wetland annual carbon sink for North America. Uncertainties in the sizes of the forested wetland soil carbon pools continue to be significant due in part to insufficient data on variabilities in carbon densities across diverse swamp types. Further, there are limited data on the rates of vertical accretion of swamp soils and the associated long-term rates of carbon accumulation, needed for better predicting impacts of climate warming on carbon sequestration in swamp soils. We present here a comparative synthesis of swamp soil carbon properties including bulk densities, organic carbon contents, soil thicknesses, rates of vertical accretion and rates of long-term carbon accumulation, from >200 swamp sites. We compare these properties for broad-leaf swamps (including mangroves), needle-leaf swamps, mixed swamps, and shrub-dominated swamps, and also compare across North American Ecoregions. The results show significant variability across peat-forming and mineral swamps, and indicate rates of carbon accumulation in some swamp types similar to those of northern bogs and fens.

Virtual networking opportunities are becoming increasingly important as alternatives or add-ons to offline conferences to lower the academic carbon footprint and make academia more inclusive. Virtual conferences improve access to networking opportunities and knowledge exchange for researchers, e.g., from the Global South or those with disabilities or caregiving obligations. Virtual conferences may be especially important for early-career researchers (ECRs), who are more often excluded from the expensive participation in offline conferences and associated networking opportunities. Our study presents the outcomes from the 2020 online edition of ICYMARE (International Conference for Young Marine Researchers, www.icymare.com), which will use an innovative approach for online networking. ICYMARE is an ideal case for this study as it is organized by and for ECRs. With around 350 participants during the on-site meeting, ICYMARE has great potential to conduct a comprehensive survey on its online alternative and its focus on marine sciences is of high relevance. Following the 2020 online edition and the outcomes of this study, the approach of this conference series may be widened to a mixed conference approach where online participation is fostered in a live conference event (e.g., through live streams, video on demand, TED talk formats). Currently, most online substitute conferences are held in various online seminar formats that focus on sharing information from keynote speakers. ICYMARE will be using a different approach, focused on networking and creating personal connections. There will be online sessions in virtual meeting rooms of around 10 participants, which include scientific presentations and moderated discussions. Additionally, an open space for general networking and free exchange is provided. As this is an innovative format, no research has been done to determine the networking potential and success of such an alternative online format and we aim to address this gap. We will use a mixed-method approach using qualitative and quantitative methods, covering aspects about the advantages, disadvantages, and challenges of online conferences for ECRs. Our presentation will provide an overview of the preliminary outcomes from this study as well as the lessons learned during this first online edition.

The ESA Virtual Space Weather Modelling Centre (VSWMC )project was defined as a long term project including different successive parts. Parts 1 and 2 were completed in the first 4-5 years and designed and developed a system that enables models and other components to be installed locally or geographically distributed and to be coupled and run remotely from the central system. A first, limited version went operational in May 2019 under the H-ESC umbrella on the ESA SSA SWE Portal. Part 3 is the next development step before all objectives of the VSWMC are achieved. The goal of the ESA project “Virtual Space Weather Modelling Centre - Part 3” (2019-2021) is to further develop the Virtual Space Weather Modelling Centre, building on the Part 2 prototype system and focusing on the interaction with the ESA SSA SWE system. The objective and scope of this new project include, apart from maintaining the current operational system, the efficient integration of new models and new model couplings, including daily automated end-to-end (Sun to Earth) simulations, the further development and wider use of the coupling toolkit and front-end GUI, making the operational system more robust and user-friendly. The VSWMC-Part 3 project started on 1 October 2019. The new models that are being integrated are Wind-predict (a global coronal model from CEA, France), the Coupled Thermosphere/Ionosphere Plasmasphere (CTIP) model, Multi-VP (another global coronal model form IRAP/CNRS, France), the BIRA Plasma sphere Model of electron density and temperatures inside and outside the plasmasphere coupled with the ionosphere (BPIM, Belgium), the SNRB (also named SNB3GEO) model for electron fluxes at geostationary orbit (covering the GOES 15 energy channels >800keV and >2MeV) and the SNGI geomagnetic indices Kp and Dst models (University of Sheffield, UK), the SPARX Solar Energetic Particles transport model (University of Central Lancashire, UK), Spenvis DICTAT tool for s/c internal charging analysis (BISA, Belgium), the Gorgon magnetosphere model (ICL, UK), and the Drag Temperature Model (DTM) and operations-focused whole atmosphere model MCM being developed in the H2020 project SWAMI. We will provide an overview of the state-of-the-art and demonstrate the system.

The increased availability of daily model output in the latest generation of climate models should allow us a greater understanding of, and an improved ability to predict, future trends in short duration extreme temperature events. We examine the changes in 1, 3, and 5 day averaged annual maximum temperature at the levels of global warming highlighted by the 2016 Paris Agreement, and an additional degree of warming. These events are characterised using daily near surface air temperature output from four large ensemble models in the SSP370 scenario of the Coupled Model Intercomparison Project Phase Six (CMIP6). Bootstrapped distributions of ensemble members are fitted to the generalised extreme value distribution and the changes in location, scale and shape parameters examined at the respective warming levels, compared to the last 10 years of historical model runs. Global trends in location parameter indicate increased warming over land relative to oceans while shape and scale parameters show less globally consistent trends but very clear signals of strong changes over the Arctic sea ice. Risk ratios are determined for extreme temperature events with a return value of 10 years in the historical period, compared to future levels of warming, using the calculated GEV distributions. Risk Ratios increase globally with mean temperature change, with greater increases over the tropics, particularly over the oceans. Longer events are also found to have a greater increase in occurrence than shorter duration temperature extremes.

SubSahara Africa has been the last continent to experience a significant number of cases in the novel Coronavirus (COVID-19). Several studies have suggested that air pollution is related to COVID-19 mortality; poor air quality has been linked to cardiovascular, cerebrovascular, and respiratory disease, which are considered co-morbidities linked to COVID-19 death. We examine potential connections between country-wide COVID-19 mortality and environmental conditions in Senegal, Cabo Verde, Nigeria, Cote D’Ivoire, and Angola. We analyze PM2.5 concentrations from cost-effective in situ measurements, aerosol optical depth (AOD), and fire count from space-borne platforms during the dry season when dust and biomass burning aerosols are present in Southern Africa. In addition, we examine the COVID-19 mortality during the wet season using space-borne rain measurements to determine potential linkages which might occur as a result of time spent indoors and concurrent cases of flu, waterborne and vector-borne diseases in West Africa. Results are presented from March through December of 2020.

Precise understanding of complex physical mechanisms of mesoscale process require high resolution observations of temperature, moisture, wind, precipitation, clouds. Using all collocated observations of microwave radiometer, wind profilers, electric field mill, weather radars over South-East India an observational analysis is conducted for the first time. Analysis suggests that these systems developed in warm, moist environment associated with large-scale low-level convergence. Passage of system is accompanied by convective regions with intense upward motion and towers extending up to higher levels indicating developing phase and presence of upward/downward motion comprising of heavy precipitation representing mature phase of the system followed by stratiform regions with prominent downdraft motion and less precipitation related to decay phase. Large (small) values of reflectivity and cloud liquid water values represent presence of deep (shallow)convective (stratiform) regions. Cloud to Ground (CG) lightning activity associated with storm electrification processes showed the existence of both +CG and –CG flashes in convective and dominant –CG in stratiform regions. Presence of different sized cloud liquid hydrometers in convective regions resulted in bipolar nature due to their collisions however in stratiform regions their distribution is mostly uniform and resulted in single polarity. Combination of different observations has provided the unique opportunity to examine interrelations of different physical mechanisms in storm environment. Inspection of reflectivity, CG lightning and cloud liquid water measurements have demonstrated the relationship of lightning mechanism with storm dynamics and cloud microphysics. Combined investigation of temperature, moisture and wind measurements have given considerable insight of thetae ridge formations resulting from thermal and moisture advections. Isentropic upgliding and downgliding facilitated the unique way to visualize the vertical transport of temperature and moisture through ascent and descent of air parcel. Blend of observations presented considerable insight of synoptic and complex mesoscale processes and their mutual interactions in the storm environment and provided encouraging results in explaining MCS structure.

Cloud-top heights (CTH) from the Multiangle Imaging Spectroradiometer (MISR) and the Moderate Resolution Imaging Spectroradiometer (MODIS) on Terra constitute our longest-running single-platform CTH record from a stable orbit. Here, we provide the first evaluation of the Terra Level 2 CTH record against collocated International Space Station Cloud-Aerosol Transport System (CATS) lidar observations between 50ºN - 50ºS. Bias and precision of Terra CTH relative to CATS, calculated from the normality of CTH error histograms, are shown to be strongly tied to cloud horizontal and vertical heterogeneity and altitude. For single-layered, unbroken, optically thick clouds observed for all altitudes, the uncertainty in MODIS and MISR CTH are -540±690 m and -280±370 m, respectively. The uncertainties are generally smaller for lower altitude clouds and larger for optically thinner clouds. For multi-layered clouds, errors are summarized herein using both absolute CTH and CATS-layer-altitude proximity to Terra CTH. We show that MISR detects the lower cloud in a two-layered system, provided top-layer optical depth < ~0.3, but MISR low-cloud errors are unaltered by the presence of thin cirrus. Systematic and random errors are propagated to explain inter-sensor disagreements, as well as to provide the first estimate of MISR stereo-opacity bias. For MISR, altitude-dependent wind-retrieval bias (-90 to -110 m) and stereo-opacity bias (-110 to -240 m) and for MODIS, bias due to low opacity near cloud-top lead to overall negative CTH bias. MISR’s precision is largely driven by wind-speed uncertainty (3.7 m s-1), whereas MODIS precision is driven by forward-modeling uncertainty.

Bangladesh, part of Indo-Gangetic-Brahmaputra Plain, is frequently affected by floods and droughts. As the economy of Bangladesh is still agriculture based, effective measurement of soil moisture will not only strengthen the irrigation management but also improve the hydrological modelling and drought prediction. But only nine agro-meteorological stations of Bangladesh measure the soil moisture four times a month which creates a vacuum to scientifically manage her water resources. SMAP (Soil Moisture Active Passive) satellite of NASA provides an unprecedented opportunity for full scale measurement of soil moisture over this region. Field measurements of soil moisture from April 2015 were used to assess the effectiveness of the SMAP’s measurement over the North West Region of Bangladesh which suffers from frequent dry spells. Initially the Root mean squared error (RMSE) between the SMAP and observed soil moisture were found to vary between 12.28 to 16.72% for the available stations. The results showed a bias in SMAP data and it was significantly reduced using bias correction. Later multiple linear regression, based on supplementary climate data in addition to SMAP observations, was applied to obtain an improved estimate of soil moisture and the RMSE were reduced to 1.19 to 3.18%. Lastly, different machine learning techniques (i.e. ANN, SVR, XGBoost etc.) were used to reduce the bias further. This study demonstrates a promising potential of using the SMAP data in soil moisture estimation over Bangladesh for its effective water resources management.

Conceptual models are among the most frequently used type of models in watershed modelling studies, due to their low computational requirements and ease of interpretation. Model selection requires the comparison of model alternatives, which is complicated when the models present in the literature differ in many uncontrolled aspects, such as conceptualization, implementation, and source code availability. To overcome this limitation, several model-building frameworks have been introduced in the last decade, which facilitate model comparisons by enabling different model alternatives within the same software and numerical architecture. Building on the decennial experience with the development and usage of Superflex, a flexible modeling framework for conceptual model building, so far implemented in FORTRAN language and not available as open source, we propose SuperflexPy, an open source Python framework for building conceptual hydrological models. Compared to other existing models or flexible frameworks, SuperflexPy is designed to be extremely easy to modify or extend, allowing scientists to build models that reflect their processes understanding; thanks to its object-oriented architecture and its complete integration with the Python programming environment, SuperflexPy can be seen as a high-level programming language for constructing hydrological models that are extremely flexible both in terms of elements configuration (i.e. how the elements are connected into a structure) and spatial organization. By design, SuperflexPy is not limited to water quantity but can be extended to simulate transport processes (water quality). In this presentation, we will illustrate the principles behind the design of the framework and showcase some applications.

One global-scale dust storm and two larger-than-average regional-scale dust storms have been recorded by spacecraft observations within the dust seasons of the past two Mars years. Each of these storms began near times when the orbit-spin coupling torques on Mars were changing most rapidly. While a clear association between Martian planet-encircling dust events and orbital torque episodes has been established (JGR-Planets 125, e2019JE006077), possible relationships between orbital/dynamical variability and regional-scale storm occurrence have not previously been investigated. In this study we compare the initiation periods and early development of the MY 34 “C” storm and the MY 35 “A” storm with prior findings obtained for the planet-encircling dust event of MY 34. We employ observations by MRO’s Mars Climate Sounder to compare dust loading, atmospheric temperatures, dynamical heating indices, and dust layer peak altitudes, to illuminate the similarities and differences between these events, in juxtaposition with a discussion of the concurrent variability of the planetary orbital angular momentum and the resulting torque on Mars.

On 16 December 2017 a chain mass movement, started as a rockslide, produced a debris and mud flow which channelized in the Burritos River. The flow reached and flooded the Santa Lucia village (-43.41°/-72.37°) destroying 50% of the urban area and killing 22 people. This fatal event resulted as a conjunction of geologic, meteorological and glaciological variables: Highly altered volcanic rocks with deep vertical fractures and steep slopes in the Yelcho Range; an intense precipitation event 30 hours prior the rockslide of up to 124.8 mm in one day with a high zero-degree isotherm elevation; an ice-cored lateral moraine deposit under the headwall plus hyper saturated soils downhill near the Burritos River. The role of a proglacial lake has been discussed, nonetheless, current studies argue that the main liquid inputs could have been pre-existing water in saturated soil. In this study we analyzed satellite imagery available on the web, including Sentinel 2 and Landsat scenes. Based on the Public Chilean Glacier Inventory, we quantify the retreat of glaciers laying on the eastern flank of the Yelcho Range, west of the Santa Lucía village. On a time-span of 33 years (1987 to 2020), we estimate over 650 m of glacial retreat plus an accelerated retreat rate up to 30 m/yr for the last decade. Such glacier shrinkage coupled with increasing local temperatures may lead to the genesis of more proglacial lakes, enhancing the odds of chain mass movements, favoring the interaction between rockslides and glacial lakes. Our results show an example of a relatively simple analysis that can be performed with basic tools and with no further expenses. The proper understanding of glacial landscape evolution with time and its response to extreme climatic events is essential to prevent human loses, as these events will be more frequent due to the current climate change context. For that purpose, further work aims to clarify more hidden variables in the region, such as the number of proglacial lakes, steep slopes on headwalls, highly eroded slopes, among others, in order to enlarge the local knowledge linked to one of many components involved on a natural disaster, thus, increasing the resilience of communities whom are still occupying the affected area after the Santa Lucía landslide disaster.

Recent increasing wildfire activities in California have raised widespread public concerns. This raises a question as to whether the increase of of fire weather this year and in recent decades is mainly due to weather pattern change dominated by natural variability or anthropogenic warming. We observed that vapor pressure deficit (VPD), an important fire weather risk index, in the warm season (May to September) of California, has increased significantly since 1979, with a positive trend of 0.70 hPa or 0.50 σ per decade, and this trend is also significant for VPD in large fire days; trend analysis of saturated and actual vapor pressure suggests that direct warming and drying contributes to about 70% and 30%, respectively, to the positive VPD trend. This increase of VPD and warming for such a short period cannot be simply linked to anthropogenic factors as it can also be affected by circulation changes. Our self-organizing maps analysis reveals the close relationship between VPD and circulation. Using a flow analogue approach, we estimated circulation changes contribute to about half (54%) of the observed VPD trend. Since circulation changes represent the upper limit of internal variability of the climate system because both the internal climate variability and anthropogenic forcings can contribute to such a change, our results suggest that anthropogenic warming must explain at least the other half of the observed VPD trend.