1. Introduction
Every summer in the Northern Hemisphere there is a distinct poleward shift in cyclone frequency from the extratropical storm tracks in the North Atlantic and Pacific Oceans to the central Arctic (e.g., Serreze and Barrett 2008). This seasonal transition is connected to the poleward migration of the Northern Hemisphere Annular Mode (Ogi et al. 2004)—the leading mode from an empirical orthogonal function analysis of the zonal mean geopotential height field (Thompson and Wallace 2000)—and the emergence of low-level baroclinicity from thermal gradients associated with the sea ice-land thaw during the Arctic melt season (late May–August). Typical forcing from large-scale baroclinic instability is not absent, but weaker in summer and therefore low-level differential heating becomes a primary catalyst for high-latitude cyclone development with the majority of cases forming over interior Eurasia or along the continent’s north coast before entering the central Arctic (Serreze and Barry 2005). These development and propagation pathways often position the warm-sector circulation for many cyclones either over or adjacent to the Eurasian landmass (Figs. S1c,g in the online supplemental material), which from July through August is typically snow free (Estilow et al. 2015) and has reached its annual maximum for moisture flux to the atmosphere (Jiménez et al. 2011). In this study, we explore this co-positioning of land surface moisture fluxes and cyclone circulation to assess whether the Eurasian landmass serves as a summer season moisture source region for cyclone uptake and moisture transport into the Arctic.
Previous studies have examined source-receptor connections between evapotranspiration (ET), precipitation P, and atmospheric circulation (including cyclones) at global and regional scales (e.g., Trenberth 1999; Brubaker et al. 1993, respectively) using the recycling ratio—a measure of the fraction of precipitation over a defined area that originated as evapotranspiration from that same area. For a region over western Eurasia (30°–60°N and west of the Ural Mountains ∼30°–50°E), Brubaker et al. (1993) estimated a summer recycling ratio of ∼31%. East of the Urals over Siberia, Serreze and Etringer (2003) determined an average recycling ratio of ∼25% for each of the three largest Eurasian–Arctic watersheds (i.e., the Ob, Yenisey, and Lena; Fig. S2 in the online supplemental material), which are similar in size geographically. The size of the analysis region is a determining factor in the recycling ratio calculation (Dirmeyer and Brubaker 2007), as other studies show that the summer recycling ratio determined for broad sectors of northern Eurasia is greater than 50% (Kurita et al. 2004) or even approaches 80% when the entire continent is considered (Koster et al. 1986; Numaguti 1999). These estimates indicate that the land surface plays an important role in the summer hydrological cycle and that a portion of the land surface moisture flux (∼20% in the case for the entire Eurasian continent) is not recycled but rather exported by atmospheric circulation.
Enhanced evapotranspiration (ET > P) from high-latitude soils during the summer months is a common result in a number of recent studies. Evidence of greater transpiration has been detected in multiyear remote sensing measurements such as surface greening in tundra regions (e.g., Berner et al. 2020) and increased forest productivity in boreal biomes (e.g., Piao et al. 2011), with the greatest sensitivity found along permafrost boundaries at Eurasian tundra sites (e.g., Myers-Smith et al. 2015). At the same time, observations and multiple reanalysis datasets show a positive trend in evaporation over a broad sector of north-central and western Eurasia (Serreze et al. 2003; Nygård et al. 2020). It has been suggested that these changes in evapotranspiration in recent decades are in response to warming of traditionally colder and wetter soils, which in turn directly augment the land–atmosphere coupling (Dirmeyer et al. 2013). For the purposes of this study, we consider these trends but examine soil conditions in the context of weather time scales. Our objective is to evaluate to what extent high-latitude soil moisture and land surface latent heat fluxes supply summer Arctic cyclone-induced poleward moisture transport.
Previous studies show that transient eddy circulations (e.g., cyclone and anticyclones) are the primary drivers of moisture transport into the Arctic (e.g., Dufour et al. 2016), with individual cyclones (poleward of 50°N) accounting for 74% of the average annual amount and greater than 80% of the average summer amount (Fearon et al. 2021). Moisture transport into the Arctic—injections of high moist enthalpy poleward of 70°N, namely, moist-air intrusions or simply intrusions (Doyle et al. 2011; Woods et al. 2013)—and the associated boundary layer clouds are capable of inducing both radiative cooling and warming (e.g., Persson et al. 2002; Shupe and Intrieri 2004; Wang et al. 2001). Intrusion airstreams, which are the focus of this study, have many similarities to atmospheric rivers in that they are transient, low-level filaments of relatively warm and moist air that are detected using thresholds on integrated vapor transport or IVT (e.g., Shields et al. 2018). In the Arctic, however, the implications of these intrusions are unique because they introduce large perturbations to the surface energy budget and can lead to transitions in the atmospheric conditions throughout the region. A number of studies have focused on winter intrusions (e.g., Woods et al. 2013), as cold-season intrusion events often result in the most distinct differences in net longwave radiation, where clear and cold conditions (net LW ∼ −40 W m−2) are replaced by a scenario that is relatively warm and cloudy (net LW ∼ 0 W m−2; Stramler et al. 2011). Similar transitions accompany intrusions in summer, but the change in longwave radiation can be less distinct since the background atmospheric state is seasonably warmer. Observations taken beneath moist-air intrusions reveal anomalous low-level moistening and warming during all seasons, which, for example, can lead to situations of sea ice melt in summer (Tjernström et al. 2015) or delayed ice growth in winter (Persson et al. 2017).
Intrusion frequency increases from multiple sectors in summer, especially from the surrounding continents (Fig. S1g in the online supplemental material; Fearon et al. 2021). Previous studies have linked summer intrusion water vapor to continental sources such as the Eurasian boreal forest (Vázquez et al. 2016) and the broader Siberian landmass (Komatsu et al. 2018). Komatsu et al. (2018) used observations and numerical experiments to show that atmospheric latent heating and cloudiness in the warm sector circulation of an Arctic cyclone were modulated by low-level moisture carried poleward from the lower latitudes. They suggested that a portion of the imported moisture was connected to continental sources over Siberia. Despite this effort, the extent to which the continental land surface supplies moisture to intrusions remains unclear.
In this study, we revisit the link between the land surface and moist intrusions in the Arctic. And while the primary moisture source for intrusions is undoubtedly from the oceans, we hypothesize that summer-season warming and thawing of traditionally cold or partially frozen high-latitude soils makes the land surface a significant moisture source. To evaluate this hypothesis, we first examine high-latitude continental moisture source regions in relation to cyclone-intrusion airstreams using multiyear reanalysis data and back trajectory calculations. We then employ the Coupled Ocean–Atmosphere Mesoscale Prediction System (COAMPS) and regional soil moisture reduction experiments to diagnose the land surface moisture contribution for an individual Arctic cyclone case in August 2016. The details of the reanalysis data and COAMPS simulations are described in section 2. Section 3 provides an overview of summer cyclone frequency, moist intrusions, and continental moisture uptake using reanalysis and back trajectory calculations. Results from the COAMPS simulations are discussed in section 4. Section 5 provides a discussion of the results and conclusions.
2. Model and data description
a. Reanalysis data
Our analysis begins with an overview of summer Arctic cyclone and moist intrusion frequencies derived from the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA-5; Hersbach et al. 2019). The ERA-5 dataset was chosen for this study to leverage previous results from Fearon et al. (2021) on moist intrusions and Arctic cyclones. ERA-5 data are also used as validation for COAMPS simulations for parameters such as mean sea level pressure and cyclone track. ERA-5 has a horizontal resolution of 0.25° × 0.25° interpolated from a native Triangular Linear (TL) Gaussian grid TL1279L137 from the surface up to 0.01 hPa. Our analysis uses pressure level data (1000–300 hPa, every 25 hPa) at hourly and 6-hourly temporal resolution for May–September from 1998 to 2017.
b. COAMPS
In the second part of our analysis, we examine an individual August 2016 Arctic cyclone case described in section 4a and evaluate the soil moisture contribution to water vapor in a cyclone-induced moist intrusion using the COAMPS (Hodur 1997; NRL 2003). In this study, all simulations are performed using a terrain-following, nonhydrostatic atmospheric component of the COAMPS with a 15-km horizontal grid mesh (477 × 477 points) centered on the North Pole and with 70 vertical levels. Forecast fields from the NOAA Global Forecast System (GFS; 0.5° × 0.5°) provide the initial and lateral-boundary conditions. One reason for this choice is that GFS fields are well established for use in the COAMPS framework. Moreover, this choice reserves ERA-5 reanalysis as an independent data source for climatological calculations, analysis, and use for evaluation against the COAMPS control simulation. Land–atmosphere processes are described by fully coupled two-way interaction, while processes for ocean and sea ice are performed uncoupled using prescribed initial conditions and one-way interaction. These coupling choices reflect our primary research objective in evaluating land–atmospheric processes, specifically the transfer of water vapor. Land surface processes in the COAMPS are simulated using version 3.2 of the Noah land surface model (Noah-LSM, see Yin et al. 2016) with four active soil layers that range in depth from the surface to 200 cm in increments of 0–10, 10–20, 20–100, and 100–200 cm. Data from the Land Information System (LIS) framework (NASA 2022) at 0.25° × 0.25° horizontal resolution are interpolated to the COAMPS grid and provide initial conditions for land surface variables that include canopy moisture, snow depth and water equivalent, and vegetation fraction/type. LIS data are also used to initialize soil moisture, water content, and temperature for all four soil layers. Other model physics and parameterizations used in COAMPS simulations follow those described in Doyle et al. (2014).
For the purposes of our cyclone case study, a COAMPS control run and two experiments with reduced regional soil moisture are performed. The simulations run for 120 h and we examine hourly output. Each simulation was initialized at 0000 UTC 11 August 2016 and used the same model physics. There are two differences between the control and the two experimental simulations. The initial soil moisture in all four soil layers is reduced by 50% and 100% in each respective experiment at all grid points within an area over northwestern Eurasia (see section 4b). Rainfall values for these same grid cells are set to zero when passed to the Noah land surface model for all time steps in both experiment simulations so that precipitation cannot increase soil moisture within this box during the simulations. In these grid cells in which rainfall was set to zero, simulated soil moisture values at all lead times in both experiments showed little or no change relative to the initial time. All other processes in each experiment simulation follow that of the control, including those that alter soil moisture—for example, evaporation or moisture transport within soil layers. In the discussion that follows, experimental simulations for soil moisture reduction by 50% and 100% are referenced by abbreviations SM = 50% and SM = 0, respectively.
Since previous studies have shown that the size of the analysis region can affect regional moisture calculations (e.g., as with recycling ratio, Dirmeyer and Brubaker 2007), we performed additional sensitivity tests in which we halved or doubled the size of the area over which soil moisture was reduced for the SM = 0 experiment. Expanding the soil moisture reduction area had almost no impact on the moist intrusions or the cyclone but reducing the size of the soil moisture reduction area resulted in higher moisture within the intrusion, and slightly more cyclone intensification. This was largely due to the cyclone circulation having greater access to higher soil moisture along the periphery of the edges of the reduced domain. These tests emphasize the localized nature of soil moisture influences on moist intrusion properties and cyclone intensification for this particular case.
While this study focuses on impacts to cyclone-induced poleward moisture transport under regional soil moisture decreases, we also perform a COAMPS simulation where regional soil moisture is increased over northwestern Eurasia in a manner consistent with the original soil moisture reduction experiments. Regional soil moisture increases were set according to the soil porosity, which is the fraction of the total soil volume that is taken up by the pore space through which moisture/water transport occurs. In COAMPS, soil porosity represents an upper limit or soil moisture saturation threshold. Results from this experiment were similar to the control simulation with some exceptions. Several quantities such as integrated vapor transport (section 2c), low-level winds, cyclone track, and cyclone central pressure exhibited low sensitivity to soil moisture increases. However, other quantities such as low-level water vapor and cloud water mixing ratios (e.g., 850 hPa and below) were slightly more sensitive. We discuss these results in section 5.
c. Description and detection of moist intrusions
d. Air parcel trajectories
To evaluate land surface moisture source regions and to quantify moisture differences within cyclone intrusions, backward trajectory calculations are performed in this study using Lagrangian Analysis Tool (LAGRANTO; Wernli and Davies 1997; Sprenger and Wernli 2015). For COAMPS simulations, 10 back trajectories are computed for each forecast hour from within the cyclone intrusion, with each trajectory’s horizontal starting location (x and y) selected based on the 10 largest values of IVT. The starting altitude z at each x and y location is then assigned based on the level of where the horizontal moisture flux qυh is largest. Trajectories are tracked in x, y, and z space, where z represents a terrain-following sigma surface, at hourly intervals backward in time to the initialization hour. Trajectory positions (all points in x, y, and z space along the trajectory path including the starting positions) computed from the control simulation were also used to derive parcel quantities in experiment simulations. This choice allows for direct intercomparison of parcel quantities (e.g., water vapor mixing ratio) across all simulations. We did examine the sensitivity of trajectory starting positions and the implications on parcel paths computed from experiment simulations relative to the control; however, positional changes were small as were the magnitude differences in parcel quantities. For ERA-5, back trajectories are computed on pressure surfaces (x, y, and p) every 6 h in a similar manner as in COAMPS, except the starting position (x and y) is selected from within moist intrusion grid cells along 70°N where IVT is largest. Trajectories are tracked for a period of 5 days—the approximate Lagrangian residence time of atmospheric water vapor (Läderach and Sodemann 2016). ERA-5 back trajectory positions detected on pressure surfaces below model terrain height are assigned a value of undefined and are not used in the analysis [see Fearon et al. (2021) for more details on ERA-5 trajectories].
3. Overview of summer cyclones, intrusions, and continental land surface fluxes
As described earlier, cyclone frequency in summer, unlike during the other seasons, increases over the eastern and central Arctic and interior Eurasia (Fig. S1c in the online supplemental material). This poleward shift from the traditional storm tracks occurs in response to an increase in surface baroclinicity from sea ice melt and land surface thaw along ∼70°N (Crawford and Serreze 2016). This shift in summer cyclone activity is also reflected in the frequency of their attendant moist-intrusion airstreams (Fig. S1g), where multiple pathways into the Arctic emerge, particularly from continental North America and Eurasia.
Coincident with this activity, latent heat fluxes from the underlying continental landmass become increasingly active in the late spring and summer months. In Fig. 1, spatial composites of latent heat fluxes, moist-intrusion airstreams (as back trajectories), and the associated cyclones are shown for May–September. During May–July, a strong increase in latent heat flux is apparent across portions of western North America and a broad swath of Eurasia. Surface fluxes decline in August and September (Figs. 1a–e). In Figs. 1f–j, back trajectories initiated from within moist-intrusion airstreams are depicted for the same period. Only land-origin trajectories—parcel paths where 75% or more points are positioned over land—are shown for altitudes below 1 km. Trajectory segments are shaded by parcel water vapor mixing ratio and indicate where parcel moistening or drying occurs as parcels travel poleward over the land surface into the Arctic. In comparing the month-to-month transition in trajectories (Figs. 1f–j) and fluxes (Figs. 1a–e), spatial and temporal consistency is apparent between parcel moisture gains (losses) and increases (decreases) in latent heat flux to the atmosphere. The corresponding cyclone frequencies that are associated with trajectories are shown in Figs. 1k–o. Differences in cyclone patterns from May to September appear modest. However, the number of cyclones increases during May–August before declining in September. In all months, cyclones are clustered over the northern interior of Eurasia, north-coast Eurasia, and the central Arctic with more spatial variability apparent over the eastern and central Arctic. During July and August, there is also a noticeable increase in cyclone frequency over Alaska and the Canadian Archipelago.
(left) Monthly averaged surface latent heat flux (W m−2; negative values represent upward fluxes from land to atmosphere), (center) 5-day back trajectories from moist intrusions that have at least 75% of points over a land surface (shaded by water vapor mixing ratio; g kg−1), and (right) cyclone-center counts for the cyclone cases corresponding to the moist intrusion trajectories. All fields are computed over 1998–2017 from 6-hourly ERA-5 reanalysis fields in (a),(f),(k) May; (b),(g),(l) June; (c),(h),(m) July; (d),(i),(n) August; and (e),(j),(o) September. The diurnal cycle was removed from monthly averaged latent heat flux values in (a)–(e) using a 24-h averaging period centered on each reanalysis time. Back trajectories were initiated at 70°N from within individual moist intrusions (one trajectory per intrusion) and represent air parcel paths below 1 km. Both cyclones and trajectories represent a subset of those shown in Figs. 1 and 2 of Fearon et al. (2021).
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
In Fig. 2, we quantify total monthly moist-intrusion flux at 70°N associated with land and nonland trajectories. Land-origin trajectories are defined as in Figs. 1f–j while nonland trajectories represent all the remaining parcel paths not shown in Figs. 1f–j, including those with all points over ocean or those with less than 75% of the points over land. Fluxes are also stratified for the Eastern and Western Hemisphere to quantify which continental region (North America versus Eurasia) represents the largest moisture source for total monthly intrusion flux. Inspecting Fig. 2, we find that fluxes peak in July for each category. The total monthly intrusion flux predominately comes from the Eastern Hemisphere, with fluxes more than double those found in the Western Hemisphere. Nonland trajectories also stand out as the primary contributor to the total intrusion flux, which is not surprising since the oceans represent a major moisture source. Looking further at land trajectories in the Eastern Hemisphere, ∼30% of the total intrusion flux during all months is associated with land regions (e.g., Eurasia). In the Western Hemisphere, significantly less of the intrusion flux (<17%) is associated with land trajectories. This result is in part due to less land area in the Western Hemisphere and less overall Arctic cyclone activity (Figs. 1k–o).
Monthly averaged intrusion IVT (kg m−1 s−1) across 70°N computed from 6-hourly ERA-5 reanalysis within moist intrusions of cyclones occurring between 1998 and 2017 in May, June (JUN), July (JUL), August (AUG), and September (SEP). The total monthly average intrusion IVT across 70°N (label T) is calculated separately for intrusion trajectories with more than 75% of points over land (“Land,” magenta bars, as shown spatially in Figs. 1f–j) and for the remaining intrusion trajectories, which we call “Non-land” (dark purple bars). Total monthly average intrusion IVT is also partitioned for the Eastern and Western Hemisphere (labels E and W, respectively). Percentages of the total monthly IVT for non-land and land categories are provided in the table along with partitions for the Eastern and Western Hemispheres, as labeled.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
Further inspection of Figs. 1 and 2 reveals that monthly averaged latent heat flux (Figs. 1a–e), trajectory mixing ratio (Figs. 1f–j), and the percentage of land intrusion flux (Table in Fig. 2) all gradually increase from May to July before decreasing in August–September. This result remains consistent when land intrusion fluxes are stratified for the Eastern and Western Hemisphere (Fig. 2). To quantify this finding further, we integrated the positive changes in both water vapor mixing ratio and latent heat flux along 6-hourly land-trajectory segments within each month (Fig. 3). Boxplot distributions clearly indicate a July maximum in both mixing ratio gains (mean values of 1.9 and 1.5 g kg−1) and upward latent heat flux (mean values of 287 and 238 W m−2) along trajectories that traverse high-latitude land surfaces prior to entering the Arctic, in the Eastern and Western Hemisphere, respectively.
Integrated positive gains in (a) water vapor mixing ratio and (b) latent heat flux along 6-hourly land trajectory segments derived from ERA-5 reanalysis over 1998–2017 for May, June, July, August, and September. Gains in mixing ratio and latent heat flux are defined as 6-hourly positive changes that occur simultaneously in both parameters along a land trajectory segment. Boxes extend from the 25th to 75th percentile, with the median and mean values indicated as an orange line and blue diamond, respectively. Whiskers identify the minimum and maximum values that represent 1.5 × interquartile range from the 25th and 75th percentiles. Outliers are shown beyond box-and-whiskers as black circles.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
This composite analysis may mask some of the important details at the mesoscale on how the land surface contributes to the development of moist intrusions during the warm season. Therefore, in the next section, we use COAMPS simulations and sensitivity tests to evaluate the extent to which the land surface contributes moisture to an Arctic cyclone-induced moist intrusion. We select an August 2016 cyclone case that originated over northwestern Eurasia for these simulations. Our reasons for choosing an August cyclone are that the summer Arctic cyclone count is largest in August, the land surface is often completely snow free by this time, and land surface latent heat fluxes continue to be substantial. We prioritized this particular August 2016 cyclone because it originated over northwestern Eurasia—a region with a documented positive trend in summer season evapotranspiration over the last two decades (Nygård et al. 2020), and where we expect soil moisture to play an important role in modulating moisture transport into the Arctic in the coming years. Although this particular cyclone case does not strongly impact sea ice, its land origin over Eurasia makes it ideal for addressing the primary study objective of understanding how soil moisture influences Arctic cyclones and their accompanying moist intrusions.
4. Arctic cyclone case study
a. Case overview
The cyclone examined in this study developed on 11 August 2016 along the Scandinavia–Russia border near the White Sea (Fig. 4d, low pressure symbol). The synoptic pattern 24–72 h prior to surface cyclone development featured a strong upper-level trough that amplified as it moved across Scandinavia and encountered an enhanced low-level enthalpy gradient positioned along the entire northern coast of Eurasia. This strong baroclinic pattern spawned a broad area of cyclone activity over the Barents Sea. During the same time, a large occluded cyclone was evolving over the central Arctic. This larger vortex remained over the Pole during the entire month of August and exhibited multiple bursts of re-intensification as smaller cyclones entered the central Arctic (including the case-study cyclone) and merged with its circulation (see also analysis by Yamagami et al. 2017).
Plan view maps for 300-hPa geopotential height (m; black contours), mean sea level pressure (MSLP; white contours) every 2 hPa from 960 to 1008 hPa, 925-hPa equivalent potential temperature (K; color bar), and the 15% sea ice concentration contour (cyan line contour) from ERA-5 reanalysis for 8–13 Aug 2016 every 24 h. A white “L” identifies the approximate location of the low pressure center for the case-study cyclone in (d)–(f).
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
An incipient surface circulation that would become the case-study cyclone first appeared in the ERA-5 MSLP field on 1200 UTC 11 August 2016. It formed in the wake of a preceding cyclone that developed nearby on 8 August (Figs. 4a–c). The case-study cyclone propagated northeastward across the Barents Sea and developed slowly with a central pressure of 992 hPa as it reached the western shore of Novaya Zemlya on 1200 UTC 13 August (Figs. 4f and 5b magenta line at 60 h). Upon crossing the northern tip of Novaya Zemlya, the cyclone deepened more rapidly reaching a central pressure of 979 hPa on 0500 UTC 14 August. During the next 10 h, the rate of intensification temporarily slowed before increasing again as the cyclone continued moving northeastward into the central Arctic and merged with the larger vortex positioned over the pole on 14–15 August (Figs. S3a–c in the online supplemental material).
(a) Top-layer (0–10 cm) soil moisture fraction (color shading) for the COAMPS case study region at the 0000 UTC 11 Aug 2016 initialization time. Cyclone track lines are overlain in solid black, dashed black, solid red, dashed red, and magenta for the COAMPS control simulation, 27-km COAMPS control simulation, SM = 50% experiment, SM = 0 experiment, and ERA-5, respectively, for the 24–96-h forecast period. Black and magenta point symbols are drawn every 12 h, with labels every 24 h, and highlight the forecast hour for ERA-5 cyclone track and intensity. The yellow boundary identifies the land region where soil moisture reduction was performed for COAMPS model experiments. (b) Cyclone mean sea level pressure (hPa) for ERA-5 and each respective COAMPS simulation with the same color symbols as in (a). Dates at 0000 UTC every 24 h in (b) provide a linkage between actual date and forecast hour in (a) and (b).
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
b. COAMPS case analysis
Figure 5a shows the cyclone track from the COAMPS control simulation (black line) and ERA-5 (magenta line). Cyclone track differences are generally small, except poleward of 70°N where COAMPS shows a track that is slightly farther west of ERA-5. Figure 5b shows the cyclone central pressure trace for ERA-5, the COAMPS control, and both experiments with reduced soil moisture over northwestern Eurasia. The cyclone in the COAMPS control simulation initially intensifies more rapidly than in ERA-5. To understand whether this is related to differences in horizontal resolution between COAMPS (15 km) and ERA-5 (∼27 km), we conducted an additional COAMPS simulation that is identical to the control, except with 27-km horizontal resolution. We find that the cyclone in the 27-km COAMPS simulation is considerably weaker than the higher resolution control prior to 70 h (Fig. 5b, dashed black line), although not quite as weak as in ERA-5. For example, cyclone central pressure at 56 h (i.e., the time of maximum deepening) is 983 hPa in the 15-km control simulation, 990 hPa in the 27-km simulation, and 994 hPa in ERA-5. Other contributors to the cyclone MSLP differences between COAMPS and ERA-5 may be related to the use of GFS for initial and lateral-boundary conditions in COAMPS, and/or the Noah land surface model and LIS input fields. As we will show, differences in soil moisture in the model can account for differences in cyclone intensity that are comparable in magnitude to the differences we observe between the COAMPS control and ERA-5. Nevertheless, we note that our goal in this study is not to exactly reproduce the storm as it was analyzed in ERA-5, but rather to examine the sensitivity to soil moisture within this set of full physics simulations that reasonably captures the evolution of the cyclone.
Cyclones also intensify in the simulations with reduced soil moisture over northwestern Eurasia; however, the deepening rates are weaker than in the control simulation (solid and dashed red lines in Fig. 5b). The largest discrepancy in MSLP exists at 60 h between the control and SM = 0, where the control shows a central pressure that is 10 hPa deeper. After 60 h, the MSLP in both experiments becomes similar to that of the control. The period from 72 to 76 h into the simulations corresponds to when the cyclone merged with the larger cyclone over the pole, as described above, which is likely why the cyclones in the experiments have a similar intensity as the cyclone in the control simulation by this time. Small differences in cyclone track exist between the control and the reduced soil moisture experiments, which are noticeable at ∼36 h and after ∼72 h (Fig. 5a).
Figure 6 displays each COAMPS simulation at the time of maximum deepening in the control simulation (Fig. 5b, black line at 56 h). There is a well-defined moist intrusion characterized by high IVT crossing 70°N in all simulations. However, relatively lower IVT is noticeable across the core of the cyclone’s moist intrusion in both reduced soil moisture experiment simulations when compared with the control (Figs. 6a–c, area within cyan contour). Further inspection of the MSLP and near-surface horizontal wind field for each simulation reveals a stronger and more compact cyclone structure in the control, with multiple closed isobars and a better defined cyclonic circulation. The cyclone circulation becomes progressively weaker as soil moisture is reduced (Figs. 6b,c). In addition, the low-level wind field of the cyclone circulation reveals that the intrusion core east of Novaya Zemlya is primarily supplied by two airstreams that emanate from the south and southwest. Air parcels from the south, as compared with those from the southwest, travel from the land region that has reduced soil moisture and have a relatively short residence time over the ocean surface before reaching the intrusion core. This distinction in the underlying surface properties (e.g., land vs ocean) explains why the greatest IVT differences are largely concentrated east of Novaya Zemlya (Figs. 6a–c, within the cyan contour).
(a) IVT (shading; kg m−1 s−1), mean sea level pressure (MSLP; white contours every 2 hPa from 980 to 1020 hPa), and horizontal wind vectors (white arrows) from the lowest model sigma level (5 m) for the cyclone case at 56 h in the COAMPS control simulation, and the simulations with soil moisture reduced by (b) 50% and (c) 100%. The cyan contour at 350 kg m−1 s−1 identifies the intrusion core region east of the cyclone center. The yellow boundary region identifies the land region where soil moisture reduction was performed for COAMPS model experiments.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
Figure 7 provides a similar perspective as Fig. 6 but for precipitable water, highlighting the vertically integrated moisture differences between the three COAMPS simulations. As shown with IVT, a well-defined moist intrusion is depicted crossing 70°N in all simulations. However, precipitable water within the intrusion filament differs in excess of 10 mm between the control and SM = 0 experiment (Fig. 7a vs Fig. 7c, within cyan contour). Significant moisture differences are noticeable along the entire intrusion filament from 75°N southward into northwestern Eurasia and collocated with the region of reduced soil moisture. Differences over land occur within the southerly flow associated with the intrusion and ahead of the cyclone’s cold frontal boundary (e.g., as inferred from the MSLP pattern). This indicates that the relatively lower IVT in the reduced soil moisture experiments (Figs. 6b,c) is not just due to weaker wind speeds within the moist intrusions, but also to a reduction in atmospheric water vapor.
As in Fig. 6, but the color shading is for precipitable water (mm). Wind vectors have been removed for clarity. The green line within the intrusion core region indicates the location of the cross section shown below in Fig. 9.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
In Fig. 8, the gridscale and convective precipitation are shown for each respective experiment at 56 h. The spatial distributions of precipitation remain similar across the simulations; however, the magnitudes are noticeably smaller as soil moisture is reduced. This is apparent in the intrusion core east of Novaya Zemlya, within the cyclone circulation, and south over the landmass where soil moisture was reduced. These differences in precipitation magnitude persist at later lead times as the cyclone moves poleward away from the landmass (not shown).
As in Figs. 6 and 7, but the color shading is for gridscale precipitation (mm; color bar). Black line contours show convective precipitation from 5 to 40 mm every 5 mm. Blue line contours indicate contours of MSLP every 2 hPa but only for values of ≤1000 hPa.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
To examine the vertical extent of atmospheric moisture reductions in the reduced soil moisture experiments, Fig. 9 shows the corresponding vertical cross sections of water vapor mixing ratio through the moist intrusion core (along the green line shown in Fig. 7). Water vapor mixing ratios in the lower troposphere decline abruptly as the soil moisture is reduced. The largest reductions relative to the control are between 2 g kg−1 (SM = 50%) and 4 g kg−1 (SM = 0), and are primarily confined below 750 hPa. In addition, as soil moisture is reduced, progressively warmer intrusion air is present through the entire column as shown by the potential temperature (Figs. 9b,c), and as a result, the low-level stratification is enhanced over the relatively cold ocean surface. The warmer and drier conditions introduced above the ocean surface layer (e.g., SM = 0 experiment, Fig. 9c) are a manifestation of advection from the land surface (Fig. 6c) but, as we discuss next, they are also in part a reflection of reduced cloud cover and greater incoming solar radiation in the intrusion core as it moves poleward.
Pressure-level (hPa) cross section along the green line in Fig. 7 of water vapor mixing ratio (g kg−1; color shading) and potential temperature (magenta line contours, every 2 K) at 56 h in (a) the control simulation and the simulations with soil moisture reduced by (b) 50% and (c) 100%. Black shading indicates terrain. The x axis in (a)–(c) shows latitude (top line) and longitude (bottom line) in degrees along the 790-km cross section. The first x-axis label is removed from (b) and (c) for clarity of reading.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
Since enhanced cloud cover and warm air advection in a moist intrusion is known to significantly alter the underlying surface energy budget, with potential implications for sea ice loss (e.g., Tjernström et al. 2015; Persson et al. 2017), we next examine how the changes to the simulated moist intrusions that result from reducing soil moisture over Eurasia affect cloud cover and the surface energy budget over the ocean. Figures 10a–c show horizontal depictions of the vertically integrated cloud water mixing ratio for each simulation at 56 h. As soil moisture is reduced, cloud water mixing ratio declines noticeably within the intrusion core, particularly east of the cyclone, and farther south ahead of the elongated cold front (Figs. 10a–c). Otherwise, the cloud water mixing ratio is similar with localized differences near the cyclone center and along the north side of the circulation.
As in Fig. 6, but for (a)–(c) vertically integrated cloud water mixing ratio (g kg−1; color shading) and (d)–(f) net shortwave radiative flux (W m−2; positive downward) for each COAMPS simulation. Thin white lines indicate contours of MSLP every 2 hPa but only for values of ≤1000 hPa. Sea ice concentration greater than 0.15 is outlined by the white and navy line contour.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
These changes in cloud cover within the moist intrusions have important implications for the surface energy budget, which includes net longwave and shortwave radiation flux, as well as the sensible and latent heat flux. Figures 10d–f show horizontal plan views at 56 h of the net shortwave radiation flux, which is the dominant term. As the soil moisture is reduced, the shortwave flux increases in the intrusion core and ahead of the cold front. In the control simulation, the regions of greater cloud water mixing ratio (Fig. 10a) are correlated with reduced shortwave flux (Fig. 10d) that arises from the cloud cover associated with the cyclone-intrusion that shields the surface from incoming shortwave radiation. The sensitivity experiments underscore the important role of the clouds (Figs. 10b,c), with increased incoming shortwave radiation due to reduced cloud shading when the soil moisture is decreased (Figs. 10e,f).
The surface energy budget results are quantified further in Fig. 11. Here, each term is spatially averaged within the intrusion core region poleward of 70°N for all nonland grid cells (e.g., region of cyan stippling in Fig. S4 in the online supplemental material). Figure 11 shows the differences in individual terms computed between the two sensitivity experiments, (Fig. 11a) SM = 50% − Control and (Fig. 11b) SM = 0 − Control, for the 44–72-h forecast period. In both experiments, the largest difference from the control is in the shortwave flux, which is more than 90 W m−2 larger at local noon (i.e., at 57 h; dashed vertical line) in the reduced soil moisture experiments than in the control. Differences in the other flux terms indicate a relatively weaker response to soil moisture reductions in both experiments, although we find slight increases in latent heat flux and longwave flux out of the surface in the reduced soil moisture experiments (negative differences in Fig. 11). Horizontal maps of surface temperature and moisture at different times (not shown) show warmer and drier conditions associated with the intrusion air in SM = 0 (due to reduced soil moisture). As warmer and drier air is advected over the ocean in the reduced soil moisture experiments, it results in a stronger air–sea vapor pressure deficit and leads to increases in upward latent heat flux from the ocean. In addition, the reduced cloud water mixing ratio (Figs. 10a–c) and precipitable water (Figs. 7a–c) within the moist intrusion in the reduced soil moisture experiments results in weaker downward longwave fluxes into the surface from the atmosphere in SM = 50% and SM = 0 than in the control. This explains why the longwave flux differences shown in Fig. 11 (red line) are negative.
Differences in the surface energy budget terms between pairs of COAMPS simulations: (a) SM = 50% minus control and (b) SM = 0 minus control. Before computing differences, fluxes in each simulation are spatially averaged over moist intrusion grid cells that have IVT ≥ 350 kg m−1 s−1 and are over the ocean (highlighted in cyan in Fig. S4 in the online supplemental material). The differences are plotted every hour from 44 to 72 h. Positive differences indicate more downward (or less upward) flux in the reduced soil moisture experiment than in the control. The total atmospheric energy flux at the surface (Fnet; black) is the sum of the shortwave (SWnet; orange) and longwave (LWnet; red) radiative fluxes and the sensible (Hs; blue) and latent (HL; green) heat fluxes (W m−2).
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
Results from our COAMPS simulations demonstrate that reducing the soil moisture yields a drier and warmer intrusion air mass, which impacts the distribution of Arctic clouds and the underlying surface energy budget. We now turn our attention to the upstream evolution of the intrusion air mass. We use back trajectory calculations to evaluate the moisture uptake in the vicinity of the intrusion in order to quantify the portion of the moisture that can be attributed to the continental land surface, specifically over northwest Eurasia in the region of the reduced soil moisture.
In Fig. 12, back trajectories are shown for the control and the two sensitivity experiments. All trajectories were initiated from within the cyclone-intrusion core during the 24–72 h forecast period and were traced back to the initial time. Trajectories are shaded by parcel water vapor mixing ratio (Figs. 12a–c), latent heat flux (Figs. 12d–f), top-layer soil moisture (Figs. 12g–i), and solar radiation (Figs. 12j–l). An initial inspection of trajectory motion reveals that all air parcels traverse an anticyclonic path as they move poleward across the land surface into the Arctic. Trajectory paths occur in two clusters, which is a manifestation of the early development period when the intrusion was composed of two separate moist airstreams prior to merging. The majority of air parcels originate and remain below 750 m (Fig. S5 in the online supplemental material). If we examine the moisture change of the parcels, the largest area of increasing water vapor is found where parcels transition northeastward (Fig. 10a). This location is also where largest water vapor differences occur between the control and sensitivity experiments (Figs. 10a–c), with maximum values of 4.5 g kg−1 (Control − SM = 50%) and 6.6 g kg−1 (Control − SM = 0). Surface latent heat fluxes to the atmosphere exceeding 350 W m−2 are collocated or slightly upstream from this location of increased water vapor (Fig. 12d), as well as larger values of top-layer soil moisture (Fig. 12g). Similar examples of parcel moistening are noticeable along trajectory paths, where spatiotemporal alignment exists between larger values of top-layer soil moisture, latent heat fluxes, and mixing ratio. In addition, the location and timing of moisture transfer from the land surface appears to be strongly coupled to the solar radiation. For example, upstream from the largest water vapor increases (Fig. 12a), latent heat fluxes to the atmosphere and incoming solar radiation at the surface are both maximized (Figs. 12d,j). This suggests that the amount of incoming solar radiation at the surface plays a key role in regulating the amount of evaporation that occurs at these locations. Thus, these regions are a key source of continental moisture for the intrusion air mass. As soil moisture is reduced (Figs. 12h,i), the latent heat flux enhancement by the solar radiation in these key source regions becomes limited (Figs. 12e,f).
Back trajectories for the 24–72-h forecast period from each respective COAMPS simulation indicated by the column headings. Trajectories are shaded by (a)–(c) water vapor mixing ratio (g kg−1), (d)–(f) latent heat flux (W m−2), (g)–(i) 0–10-cm soil moisture (fractional units), and (j)–(l) surface shortwave flux (W m−2). The trajectories in (g)–(i) are white wherever the soil moisture is zero or undefined (which includes ocean points). A total of 10 trajectories were initiated for each forecast hour from within the moist intrusion at locations and altitudes of the largest IVT and υqh, respectively.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
In Fig. 13, we quantify moistening from the land surface for both water vapor mixing ratio and latent heat flux along all trajectory paths for each COAMPS simulation. This calculation represents the integrated positive changes that occur simultaneously in both parameters along 1-hourly land-trajectory segments. In the control simulation, the mean integrated mixing ratio is 3.1 g kg−1 (Fig. 13a, blue diamond). We define this value as our baseline gain in water vapor from the land surface along a 1-hourly segment as air parcels move poleward. For SM = 50% and SM = 0, the mean mixing ratio values are 2.1 and 1.7 g kg−1 (Fig. 13a, blue diamonds). If we compare these values with the baseline and compute the percent change, we find an average decrease of 32% and 45% in total water vapor supplied to intrusion air parcels as they move along a land-based segment for each respective experiment. If we perform the same calculation for latent heat flux (Fig. 13b), with a baseline flux gain of 465 W m−2 (Fig. 13b, blue diamond in the control simulation), the corresponding average percent change is a decrease of 47% and 91% relative to the control in surface latent heat flux supplied to intrusion air parcels as they move along a land-based segment for each respective experiment.
Boxplot distributions of integrated positive gains in (a) water vapor mixing ratio and (b) latent heat flux along trajectory paths for each COAMPS simulation. Gains in mixing ratio and latent heat flux are defined as 1-hourly positive changes that occur simultaneously in both parameters along a land trajectory segment. Boxes extend from the 25th to 75th percentile, with the median and mean values indicated as an orange line and blue diamond, respectively. Whiskers identify the minimum and maximum values that represent the 1.5 × interquartile range from the 25th and 75th percentiles. Outliers are shown beyond box-and-whiskers as black circles.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
To further quantify the moisture contribution from the land surface, in Fig. 14 we compute the percent changes in water vapor mixing ratio along trajectory paths for each COAMPS experiment. Unlike the values shown in Fig. 13, which represent the integrated positive gains along the trajectories, this calculation represents the percent change in mixing ratio between the earliest land point along the trajectory (i.e., the land-origin point) and the northernmost land point before air parcels cross over the ocean. Air parcels generally remain below 750 m while over land (Fig. S5 in the online supplemental material), and therefore changes in mixing ratio due to changes in trajectory altitude are generally small. In the control simulation, the mean percent change in mixing ratio is 30%. We define this value as our baseline percent change in water vapor supplied to the intrusion air from the land surface as it moves poleward. In the sensitivity experiments, the percent change is reduced to 24% for SM = 50% and to −1% for SM = 0. These percentages indicate how mean water vapor flux from the land to the atmosphere declines as soil moisture is reduced. In the SM = 0 experiment, a negative percentage is consistent with a downward water vapor flux from the atmosphere to the land, which occurs through condensation and precipitation.
Boxplot distributions of percent change in water vapor mixing ratio along trajectory paths for each COAMPS simulation. The percent change calculation represents the difference between the land origin point and the northernmost land point. The difference is then normalized by the land origin point value. A positive or negative value represents an overall increase or decrease, respectively, in mixing ratio along a trajectory path. Boxplot definition and symbols are as in Fig. 13.
Citation: Monthly Weather Review 151, 7; 10.1175/MWR-D-22-0264.1
5. Discussion and conclusions
Moisture transport into the Arctic has received considerable attention in the literature, particularly during the winter months (e.g., Woods et al. 2013), as it has been shown to be a major contributor to the regional moistening trend observed in recent decades (Rinke et al. 2019). In winter, additional water vapor transported into the relatively cold, dry, and dark Arctic atmosphere can enhance the cloud cover, downward longwave radiation, and surface warming (e.g., Persson et al. 2017). In summer, the surface energy balance in the Arctic is less sensitive to intrusion episodes as conditions are warmer and cloudier (Stramler et al. 2011). However, relative to other seasons, moist intrusion frequency is higher during summer, which is partly due to an increase in the number of Arctic cyclones originating over the Eurasian landmass (Fearon et al. 2021). Moist intrusions during the summer months have been observed to enhance sea ice melt by increasing the longwave radiative flux at the surface (Tjernström et al. 2015) as well as the downward sensible heat flux (Stern et al. 2020). Previous studies have also linked summer intrusion water vapor to continental moisture sources such as the Eurasian boreal forest (Vázquez et al. 2016) and land sectors over Siberia (Komatsu et al. 2018).
In this study, the goal was to build on the previous literature and specifically evaluate to what extent the high-latitude land surface represents a moisture source for water vapor associated with summer Arctic cyclone-induced intrusions. To accomplish this, we used multiyear reanalysis to examine monthly spatial composites of continental surface latent heat fluxes along with land-origin intrusion air parcels (or back trajectories). In addition, we analyzed high-resolution COAMPS simulations of an August 2016 Arctic cyclone case and performed sensitivity experiments that quantify the evaporative contribution of soil moisture to water vapor within the attendant moist intrusion.
Results from ERA-5 reanalysis indicate that the high-latitude soil moisture is an important moisture source for moist intrusions associated with summer Arctic cyclones. The majority of summer cyclones cases form over the northern interior and coastal portions of Eurasia. This positions the warm-sector circulation adjacent to or over the landmass while the cyclone intensifies. As the warm sector circulation matures, intrusion air parcels move poleward across the continental landmass where the upward surface latent heat fluxes are at or near their annual maximum. During June–August, the average surface latent heat flux accumulated along a trajectory within an intrusion that traverses across the Eurasian landmass is 270, 287, and 225 W m−2, respectively (e.g., Fig. 3b, blue diamonds, Eastern Hemisphere). These flux values correspond to air parcel water vapor mixing ratio gains of 1.7, 1.8, 1.5 g kg−1, respectively (Fig. 3a, blue diamonds, Eastern Hemisphere). Reanalysis trajectory and flux calculations also show that moisture evaporated from land regions account for more than one-third of the total poleward atmospheric moisture flux into the Arctic during summer.
Results from our COAMPS sensitivity experiments showed that the Arctic cyclone intensification rate and associated intrusion water vapor are both highly sensitive to soil moisture. Under soil moisture reduction, the simulated MSLP field showed a cyclone surface circulation that was less well-defined. Cyclone central pressure in the early stages was substantially weaker, by as much as 10 hPa in the SM = 0 experiment. The sensitivity experiments also revealed that weaker surface latent heat fluxes are primarily due to reduced soil moisture. And while, land heat flux is function of both surface moisture and wind speed, simulations showed that wind speeds over land (where soil moisture was reduced) were not appreciably different across the simulations (Figs. S6a–c in the online supplemental material). Therefore, based on these findings, we hypothesize that progressively weaker cyclone intensification in the SM = 50% and SM = 0 experiments may be due to weaker diabatic heating, which, in this cyclone case, is restricted because of reduced latent heat flux from the land surface.
In the above hypothesis, the high-latitude land surface moisture is analogous to the Arctic Frontal Zone (AFZ). The AFZ has been described as a seasonal lower tropospheric thermal boundary and an area of baroclinicity that emerges along the Arctic continental coastlines (Serreze et al. 2001). Crawford and Serreze (2016) showed that the AFZ is not necessarily an area of cyclone development, but rather an area where enhanced cyclone intensification occurs. Other proposed summer-season contributors of low-level baroclinicity in the Arctic include thermal differences along the sea ice margin (Inoue and Hori 2011) or enhanced vertical enthalpy fluxes from the open ocean (e.g., Simmonds and Keay 2009). However, we hypothesize that latent heat fluxes from high-latitude continents can substantially contribute to intensifying Arctic cyclones by moistening inflowing air parcels and enhancing diabatic heating within the cyclone and its warm conveyer belt.
While we did not explore this hypothesis in detail, supplemental analysis of diabatic heating, potential vorticity (PV), and MSLP fields viewed at the early and later stages of cyclone development for each respective COAMPS simulation does offer support for this claim. Under soil moisture reduction, low-level diabatic heating and associated PV generation (e.g., at 850 and 925 hPa, respectively; see Figs. S7a–c in the online supplemental material), which are otherwise initiated from land surface latent heat fluxes and entrained into the cyclone circulation, are restricted and hence cyclone intensification is suppressed. Alternatively, if soil moisture is increased over the land surface, we find that changes to cyclone intensification were not significant, which suggests that the sensitivity of cyclone intensification to soil moisture and land surface latent heat fluxes becomes more influential under drier soil conditions and/or as land-to-atmosphere moisture transfer is limited.
Results from our COAMPS simulations also reinforce the notion that intrusion water vapor is highly sensitive to reduced soil moisture. Back trajectories identify key source regions over the land surface where latent heat fluxes contribute moisture to the poleward intrusion airstream. The majority of air parcels within the moist intrusion originate and remain below 750 m. The location and timing of moisture flux from the land surface is found to be strongly coupled to the incoming solar radiation, where the largest latent heat fluxes coincide with times when incoming solar radiation at the surface is maximized. The COAMPS simulations show that on average cyclone-intrusion water vapor increases by 30% as air parcels traverse the high-latitude land surface regions. In the sensitivity experiments, this percentage is reduced to 24% for SM = 50% and becomes negative for SM = 0, signifying a moisture flux that is directed toward the land surface.
The reduction in soil moisture results in a drier and warmer intrusion air mass entering the Arctic, particularly in the intrusion core region. These conditions impact the intrusion cloud field and the underlying surface energy budget poleward of 70°N. The largest impacts to the surface energy budget were in the shortwave flux. Reducing soil moisture resulted in an increase in the shortwave flux into the ocean surface of more than 90 W m−2 within the intrusion core. The cloud shading effect of Arctic cyclones has been shown to temporarily slow the rate of sea ice loss, especially in the melt season (Finocchio et al. 2020; Schreiber and Serreze 2020). For this cyclone case, we find that this cloud shading effect is reduced when soil moisture becomes limited over high-latitude landmasses.
Although our focus in this study was on the impacts of reduced soil moisture, we did investigate the consequences for regional soil moisture increases (see section 2b for experiment details). In contrast to the results from reduction experiments, regional increases in soil moisture showed isolated areas of increased low-level water vapor and cloud water (e.g., 850 hPa and below) upstream over the land surface and in the intrusion core poleward of 70°N. The moisture increases in the intrusion core also translated to changes in the underlying surface energy budget, with a decrease in the shortwave flux of up to 60 W m−2. In this case, this result suggests that the cloud shading effect is increased when the soil moisture fraction approaches saturation (or the soil porosity limit) over high-latitude landmasses.
The results of this study imply that soil moisture could modulate the extent to which cyclones are able to alter the rate of sea ice melt during the summer, at least on short time scales. And while the moist intrusion analyzed in this study did not move poleward enough to substantially impact sea ice, we believe these results, which show that cyclone-intrusion clouds and the underlying surface energy exhibit strong sensitivity to changes in soil moisture, motivate further investigation of other cyclone cases with moist intrusions that impact sea ice. As such, future studies may be able to establish a link between soil moisture and sea ice loss through cloud processes by examining a larger number of cyclone cases occurring at different times during the melt season. A future study could also revisit the sensitivity of cyclone intensification to soil moisture (and diabatic heating). Instances of enhanced intensification, prompted by land surface latent fluxes for example, may produce stronger winds and facilitate mechanical breakup of the sea ice along the ice edge.
These results also have important implications for cyclone impacts on sea ice in the future Arctic. Under most climate projections, a broader area of the Eurasian landmass is expected to be unfrozen for a longer time during the warm season (e.g., Dirmeyer et al. 2013), lengthening the window for soil moisture variability to impact cyclones and their attendant moist intrusions into the Arctic through surface latent heat fluxes. Therefore, the connection between soil moisture and sea ice variability will likely become stronger in the future Arctic, as will the need to more accurately represent land surface processes and their coupling to the atmosphere in Earth system models.
Acknowledgments.
We are grateful to the two anonymous reviewers whose comments and suggestions greatly improved this paper. We also thank Michael Sprenger at ETH-Zurich for generating the cyclone tracks from the ERA-5 reanalysis data, David Ryglicki for assistance in preparing ERA-5 data used in this study, and Erin Gleason for assistance with graphics. This research is supported by the Office of Naval Research Arctic Cyclones Departmental Research Initiative (Program Element 0601153N). A portion of this research was performed while the first author held an NRC Research Associateship award at the Marine Meteorology Division of the U.S. Naval Research Laboratory.
Data availability statement.
ERA-5 reanalysis data used in this study are freely available on public repositories, except for the filtered cyclone database and trajectory calculations, which are both available upon request from the authors. ERA-5 data are available on the Copernicus Climate Change Service Climate Data Store (https://cds.climate.copernicus.eu). Because of confidentiality agreements, supporting data from COAMPS simulations used in this study can only be made available to bona fide researchers subject to a nondisclosure agreement. Details of the data and how to request access are available from the lead author at the U.S. Naval Research Laboratory.
REFERENCES
Berner, L. T., and Coauthors, 2020: Summer warming explains widespread but not uniform greening in the Arctic tundra biome. Nat. Commun., 11, 4621, https://doi.org/10.1038/s41467-020-18479-5.
Brubaker, K. L., D. Entekhabi, and P. S. Eagleson, 1993: Estimation of continental precipitation recycling. J. Climate, 6, 1077–1089, https://doi.org/10.1175/1520-0442(1993)006<1077:EOCPR>2.0.CO;2.
Crawford, A. D., and M. C. Serreze, 2016: Does the summer Arctic frontal zone influence Arctic Ocean cyclone activity? J. Climate, 29, 4977–4993, https://doi.org/10.1175/JCLI-D-15-0755.1.
Dirmeyer, P. A., and K. L. Brubaker, 2007: Characterization of the global hydrologic cycle from a back-trajectory analysis of atmospheric water vapor. J. Hydrometeor., 8, 20–37, https://doi.org/10.1175/JHM557.1.
Dirmeyer, P. A., Y. Jin, B. Singh, and X. Yan, 2013: Trends in land–atmosphere interactions from CMIP5 simulations. J. Hydrometeor., 14, 829–849, https://doi.org/10.1175/JHM-D-12-0107.1.
Doyle, J. D., C. Amerault, C. A. Reynolds, and P. A. Reinecke, 2014: Initial condition sensitivity and predictability of a severe extratropical cyclone using a moist adjoint. Mon. Wea. Rev., 142, 320–342, https://doi.org/10.1175/MWR-D-13-00201.1.
Doyle, J. G., G. Lesins, C. P. Thackray, C. Perro, G. J. Nott, T. J. Duck, R. Damoah, and J. R. Drummond, 2011: Water vapor intrusions into the High Arctic during winter. Geophys. Res. Lett., 38, L12806, https://doi.org/10.1029/2011GL047493.
Dufour, A., O. Zolina, and S. K. Gulev, 2016: Atmospheric moisture transport to the Arctic: Assessment of reanalyses and analysis of transport components. J. Climate, 29, 5061–5081, https://doi.org/10.1175/JCLI-D-15-0559.1.
Estilow, T. W., A. H. Young, and D. A. Robinson, 2015: A long-term Northern Hemisphere snow cover extent data record for climate studies and monitoring. Earth Syst. Sci. Data, 7, 137–142, https://doi.org/10.5194/essd-7-137-2015.
Fearon, M. G., J. D. Doyle, D. R. Ryglicki, P. M. Finocchio, and M. Sprenger, 2021: The role of cyclones in moisture transport into the Arctic. Geophys. Res. Lett., 48, e2020GL090353, https://doi.org/10.1029/2020GL090353.
Finocchio, P. M., J. D. Doyle, D. P. Stern, and M. G. Fearon, 2020: Short-term impacts of Arctic summer cyclones on sea ice extent in the marginal ice zone. Geophys. Res. Lett., 47, e2020GL088338, https://doi.org/10.1029/2020GL088338.
Guan, B., and D. E. Waliser, 2015: Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. J. Geophys. Res. Atmos., 120, 12 514–12 535, https://doi.org/10.1002/2015JD024257.
Hersbach, H., and Coauthors, 2019: Global reanalysis: Goodbye ERA-Interim, hello ERA5. ECMWF Newsletter, No. 159, ECMWF, Reading, United Kingdom, 17–24, https://www.ecmwf.int/sites/default/files/elibrary/042019/19001-newsletter-no-159-spring-2019_1.pdf.
Hodur, R. M., 1997: The Naval Research Laboratory’s Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS). Mon. Wea. Rev., 125, 1414–1430, https://doi.org/10.1175/1520-0493(1997)125<1414:TNRLSC>2.0.CO;2.
Inoue, J., and M. E. Hori, 2011: Arctic cyclogenesis at the marginal ice zone: A contributory mechanism for the temperature amplification? Geophys. Res. Lett., 38, L12502, https://doi.org/10.1029/2011GL047696.
Jiménez, C., and Coauthors, 2011: Global intercomparison of 12 land surface heat flux estimates. J. Geophys. Res., 116, D02102, https://doi.org/10.1029/2010JD014545.
Komatsu, K. K., V. A. Alexeev, I. A. Repina, and Y. Tachibana, 2018: Poleward upgliding Siberian atmospheric rivers over sea ice heat up Arctic upper air. Sci. Rep., 8, 2872, https://doi.org/10.1038/s41598-018-21159-6.
Koster, R., J. Jouzel, R. Suozzo, G. Russell, D. Rind, and P. Eaglesonl, 1986: Global sources of local precipitation as determined by the NASA/GISS GCM. Geophys. Res. Lett., 13, 121–124, https://doi.org/10.1029/GL013i002p00121.
Kurita, N., N. Yoshida, G. Inoue, and E. A. Chayanova, 2004: Modern isotope climatology of Russia: A first assessment. J. Geophys. Res., 109, D03102, https://doi.org/10.1029/2003JD003404.
Läderach, A., and H. Sodemann, 2016: A revised picture of the atmospheric moisture residence time. Geophys. Res. Lett., 43, 924–933, https://doi.org/10.1002/2015GL067449.
Myers-Smith, I. H., and Coauthors, 2015: Climate sensitivity of shrub growth across the tundra biome. Nat. Climate Change, 5, 887–891, https://doi.org/10.1038/nclimate2697.
NASA, 2022: LIS Framework Land Information System. NASA, accessed 19 September 2022, https://lis.gsfc.nasa.gov/.
NRL, 2003: COAMPS version 3 description: General theory and equations. NRL Tech. Rep., 148 pp.
Numaguti, A., 1999: Origin and recycling processes of precipitating water over the Eurasian continent: Experiments using an atmospheric general circulation model. J. Geophys. Res., 104, 1957–1972, https://doi.org/10.1029/1998JD200026.
Nygård, T., T. Naakka, and T. Vihma, 2020: Horizontal moisture transport dominates the regional moistening patterns in the Arctic. J. Climate, 33, 6793–6807, https://doi.org/10.1175/JCLI-D-19-0891.1.
Ogi, M., K. Yamazaki, and Y. Tachibana, 2004: The summertime annular mode in the Northern Hemisphere and its linkage to the winter mode. J. Geophys. Res., 109, D20114, https://doi.org/10.1029/2004JD004514.
Persson, P. O. G., C. W. Fairall, E. L Andreas, P. S. Guest, and D. K. Perovich, 2002: Measurements near the atmospheric surface flux group tower at SHEBA: Near-surface conditions and surface energy budget. J. Geophys. Res., 107, 8045, https://doi.org/10.1029/2000JC000705.
Persson, P. O. G., M. D. Shupe, D. Perovich, and A. Solomon, 2017: Linking atmospheric synoptic transport, cloud phase, surface energy fluxes, and sea-ice growth: Observations of midwinter SHEBA conditions. Climate Dyn., 49, 1341–1364, https://doi.org/10.1007/s00382-016-3383-1.
Piao, S., X. Wang, P. Ciais, B. Zhu, T. Wang, and J. Liu, 2011: Changes in satellite-derived vegetation growth trend in temperate and boreal Eurasia from 1982 to 2006. Global Change Biol., 17, 3228–3239, https://doi.org/10.1111/j.1365-2486.2011.02419.x.
Rinke, A., and Coauthors, 2019: Trends of vertically integrated water vapor over the Arctic during 1979–2016: Consistent moistening all over? J. Climate, 32, 6097–6116, https://doi.org/10.1175/JCLI-D-19-0092.1.
Schreiber, E. A. P., and M. C. Serreze, 2020: Impacts of synoptic-scale cyclones on Arctic sea-ice concentration: A systematic analysis. Ann. Glaciol., 61, 139–153, https://doi.org/10.1017/aog.2020.23.
Serreze, M. C., and A. J. Etringer, 2003: Precipitation characteristics of the Eurasian Arctic drainage system. Int. J. Climatol., 23, 1267–1291, https://doi.org/10.1002/joc.941.
Serreze, M. C., and R. G. Barry, 2005: The Arctic Climate System. Cambridge University Press, 404 pp.
Serreze, M. C., and A. P. Barrett, 2008: The summer cyclone maximum over the central Arctic Ocean. J. Climate, 21, 1048–1065, https://doi.org/10.1175/2007JCLI1810.1.
Serreze, M. C., A. H. Lynch, and M. P. Clark, 2001: The Arctic frontal zone as seen in the NCEP–NCAR reanalysis. J. Climate, 14, 1550–1567, https://doi.org/10.1175/1520-0442(2001)014<1550:TAFZAS>2.0.CO;2.
Serreze, M. C., D. H. Bromwich, M. P. Clark, A. J. Etringer, T. Zhang, and R. Lammers, 2003: Large-scale hydro-climatology of the terrestrial Arctic drainage system. J. Geophys. Res., 108, 8160, https://doi.org/10.1029/2001JD000919.
Shields, C. A., and Coauthors, 2018: Atmospheric River Tracking Method Intercomparison Project (ARTMIP): Project goals and experimental design. Geosci. Model Dev., 11, 2455–2474, https://doi.org/10.5194/gmd-11-2455-2018.
Shupe, M. D., and J. M. Intrieri, 2004: Cloud radiative forcing of the Arctic surface: The influence of cloud properties, surface albedo, and solar zenith angle. J. Climate, 17, 616–628, https://doi.org/10.1175/1520-0442(2004)017<0616:CRFOTA>2.0.CO;2.
Simmonds, I., and K. Keay, 2009: Extraordinary September Arctic Sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, https://doi.org/10.1029/2009GL039810.
Sprenger, M., and H. Wernli, 2015: The LAGRANTO Lagrangian analysis tool—version 2.0. Geosci. Model Dev., 8, 2569–2586, https://doi.org/10.5194/gmd-8-2569-2015.
Sprenger, M., and Coauthors, 2017: Global climatologies of Eulerian and Lagrangian flow features based on ERA-Interim. Bull. Amer. Meteor. Soc., 98, 1739–1748, https://doi.org/10.1175/BAMS-D-15-00299.1.
Stern, D. P., J. D. Doyle, N. P. Barton, P. M. Finocchio, W. A. Komaromi, and E. J. Metzger, 2020: The impact of an intense cyclone on short-term sea ice loss in a fully coupled atmosphere-ocean-ice model. Geophys. Res. Lett., 47, e2019GL085580, https://doi.org/10.1029/2019GL085580.
Stramler, K., A. D. Del Genio, and W. B. Rossow, 2011: Synoptically driven Arctic winter states. J. Climate, 24, 1747–1762, https://doi.org/10.1175/2010JCLI3817.1.
Thompson, D. W. J., and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13, 1000–1016, https://doi.org/10.1175/1520-0442(2000)013<1000:AMITEC>2.0.CO;2.
Tjernström, M., and Coauthors, 2015: Warm-air advection, air mass transformation and fog causes rapid ice melt. Geophys. Res. Lett., 42, 5594–5602, https://doi.org/10.1002/2015GL064373.
Trenberth, K. E., 1999: Atmospheric moisture recycling: Role of advection and local evaporation. J. Climate, 12, 1368–1381, https://doi.org/10.1175/1520-0442(1999)012<1368:AMRROA>2.0.CO;2.
Vázquez, M., R. Nieto, A. Drumond, and L. Gimeno, 2016: Moisture transport into the Arctic: Source-receptor relationships and the roles of atmospheric circulation and evaporation. J. Geophys. Res. Atmos., 121, 13 493–13 509, https://doi.org/10.1002/2016JD025400.
Wang, S., Q. Wang, R. E. Jordan, and P. O. G. Persson, 2001: Interactions among longwave radiation of clouds, turbulence, and snow surface temperature in the Arctic: A model sensitivity study. J. Geophys. Res., 106, 15 323–15 333, https://doi.org/10.1029/2000JD900358.
Wernli, H., and H. C. Davies, 1997: A Lagrangian-based analysis of extratropical cyclones. I: The method and some applications. Quart. J. Roy. Meteor. Soc., 123, 467–489, https://doi.org/10.1002/qj.49712353811.
Wernli, H., and C. Schwierz, 2006: Surface cyclones in the ERA-40 dataset (1958–2001). Part I: Novel identification method and global climatology. J. Atmos. Sci., 63, 2486–2507, https://doi.org/10.1175/JAS3766.1.
Woods, C., R. Caballero, and G. Svensson, 2013: Large-scale circulation associated with moisture intrusions into the Arctic during winter. Geophys. Res. Lett., 40, 4717–4721, https://doi.org/10.1002/grl.50912.
Yamagami, A., M. Matsueda, and H. L. Tanaka, 2017: Extreme Arctic cyclone in August 2016. Atmos. Sci. Lett., 18, 307–314, https://doi.org/10.1002/asl.757.
Yin, J., X. Zhan, Y. Zheng, C. R. Hain, M. Ek, J. Wen, L. Fang, and J. Liu, 2016: Improving Noah land surface model performance using near real time surface albedo and green vegetation fraction. Agric. For. Meteor., 218–219, 171–183, https://doi.org/10.1016/j.agrformet.2015.12.001.
