1. Introduction
There has been growing interest in investigating stratospheric turbulence amongst scientists and engineers to study atmospheric dynamical processes, from improving atmospheric circulation models to better predicting and forecasting clear-air turbulence events for high-altitude aircraft and other flight missions. Stratospheric turbulence is thought to largely originate from the breaking of upward propagating gravity waves. Gravity waves are produced by strong meteorological forcing in the lower atmosphere, such as from deep cloud convection, jet stream-generated shear, or strong low-level winds passing over high terrain (e.g., [1,2,3]). As these waves move upward in the atmosphere, they grow in amplitude until they reach a point of instability, and then break and transfer energy into smaller scales, i.e., turbulence (e.g., [4,5,6]). Gravity waves commonly occur in the lee of major mountain ranges (e.g., the South American Andes, the USA Rocky Mountains, and the European Alps) and over isolated mountainous islands or peninsulas (e.g., South Georgia Island, New Zealand, and the Antarctic Peninsula) or in places where large-scale air masses of different density and moisture collide to produce energetic convective storms that overshoot the tropopause (e.g., the USA Southern Great Plains, tropical Africa, and the Tibetan Plateau).
A powerful tracer for large-scale air mass transport, including gravity waves and other stratospheric–tropospheric exchange (STE) processes, comes from cosmogenic radioactive isotopes [7,8,9,10]. Established tracers include chlorine-36 (36Cl), carbon-14 (14C), aluminum-26 (26Al), tritium (3H), beryllium-10 (10Be), and beryllium-7 (7Be) [8]. 10Be and 7Be are particularly useful because they have half-lives long enough to be of use in tracer studies of large-scale atmospheric motion (e.g., [8,11]), and because they are isotopes of the same element such that they have identical chemical properties. Furthermore, 10Be, and more recently 7Be [12], are measured using accelerator mass spectrometry (AMS), which allows for concentration measurements at much lower amounts than generally possible by radioactivity counting.
10Be and 7Be are produced mostly in the stratosphere from cosmic ray collisions with oxygen and nitrogen and then strongly adsorb to ambient aerosols [7,13,14]. Aerosols accumulate these isotopes throughout their stratospheric residency until they are transported to the troposphere where they are quickly removed by scavenging via precipitation (wet deposition) (Figure 1) or to a lesser extent dry deposition. The short half-life of 7Be (53 days) with respect to 10Be (1.4 Myr) makes the ratio of aerosol-borne 10Be/7Be highly sensitive to stratospheric residence time and results in significant latitudinal and altitude variations. Predicted 10Be/7Be ratios in the midlatitudes are around ~0.6 in the troposphere, with little to no stratospheric influence, and are near to the tropospheric production ratio [15]. Ratios of ~1.3 to 3.0 are predicted in the lower stratosphere with little to no tropospheric influence. Values of ~3.0 or higher are predicted in the altitudes above the lower stratosphere [11,16].
Estimates for how 10Be/7Be varies with altitude largely come from airplane collection campaigns (such as those compiled by [18]) or are based on the assumptions about production and decay rates. Measurement data are largely limited to the troposphere and up to a few kilometers above the tropopause (nominally between 12 and 16 km in the midlatitudes). This height limit has traditionally been constrained by the altitude at which most research planes can fly. In the extreme upper atmosphere (thermosphere), a few records of very high altitude 7Be measurements come from the surface of recovered satellites (e.g., Long Duration Exposure Facility (LDEF)) which have led to estimates of concentration of 7Be atoms at the LDEF orbital altitude [10]. However, direct observations of 7Be, and especially 10Be, are lacking in the stratosphere. High-altitude balloons routinely reach heights approaching 30 km or more (i.e., the mid-stratosphere) and offer a platform for reaching new beryllium collection heights.
The Beryllium Isotopes for Resolving Dynamics in the Stratosphere (BIRDIES) experiment was planned to target deep convection storm events formed over the Southern Great Plains, USA, using high-altitude balloons. This region is known as a global “hotspot” for midlatitude convection overshooting events [19]. The campaign was designed to test the hypothesis that troposphere-to-stratosphere intrusions during cloud overshooting events produce patterns of 10Be/7Be that relate to the extent, magnitude, and mechanism of the disturbance, and that these patterns can be measured by collecting either aerosols in situ in the stratosphere or by collecting them at the ground surface via wet deposition. We base our hypothesis on the theory that as air mixes between regions of differing age (e.g., the stratosphere and troposphere), the resulting air mass will have an intermediate ratio that reflects the relative influence of the source regions [18].
BIRDIES leveraged three other field campaigns and deployments in the region. The first was the NASA Dynamics and Chemistry of the Summer Stratosphere (DCOTSS,
This paper largely serves as an overview of the field experiment, payload designs, and high-altitude flight performance. We also highlight initial observational and model findings, with a focus on trying to tie the results back to our scientific hypotheses.
2. Materials and Methods
2.1. Payloads
2.1.1. Fluctuations in the Stratosphere Using Hot-Wire (FiSH)
FiSH is a custom-made payload designed to measure in situ velocity, temperature, and acoustic fluctuations in the stratosphere [27,28]. It was designed, fabricated, and tested by the High-Speed Aerodynamic and Propulsion Laboratory at the University of Maryland (UMD), College Park. FiSH consists of a suite of sensors, data acquisition system (DAQ), and power supply housed in a polypropylene shelter. The shelter package was designed to have a streamline shape and insulation to maintain a working temperature environment for the electronics housed. In addition, a large vertical stabilizer is extended from the body to keep the payload from excessive spinning motion. The payload weight is approximately 2.6 kg. Photographs of FiSH are shown in Figure 2.
The sensor suite includes two TSI 1210-T1.5 tungsten platinum-coated wire probes with a diameter of 3.8 μm and a length of 1.27 mm for the sensing element that are connected to a constant voltage anemometer (CVA) with a frequency response of 50 kHz. One probe works as a hot-wire and is used to measure velocity fluctuations, while the second probe acts as a cold-wire and is used to measure temperature fluctuations. Other FiSH sensors include two PCB 378A04 microphones used to measure pressure fluctuations, an Edevco 8515C-15 pressure transducer to measure ambient pressure, and two K-type thermocouples are mounted externally to measure the ambient temperature during flight. The measurements are recorded by a DTS SLICE MICRO DAQ system (Diversified Technical Systems, Seal Beach, CA, USA). The DAQ and each sensor channel are set up through the DTS SLICEWare software (version 1.08.0876). This system allows for a sample rate of up to 500 kHz per channel; however, BIRDIES used a sample rate of 40 kHz per channel. This gives a vertical spatial resolution of ~0.14 mm given an average balloon ascent rate of 5.5 m/s while allowing for sufficient memory to record the entire flight duration. Note that turbulence measurements are collected only on the ascent phase of the flight. The FiSH payload is designed to be recovered at the end of each flight.
In situ stratospheric observations of turbulence have been successfully attempted by other research groups. One of the first includes an early design by [29], who flew a sensitive propeller anemometer below a superpressure balloon to capture intermittent turbulence events in the lower stratosphere associated with shearing instabilities. More recent payloads include those developed by the Hypersonic Flight in the Turbulent Stratosphere (HYFLITS) research group and the Leibniz-Institute of Atmospheric Physics (IAP). IAP has developed and flown a payload called the Leibniz-Institute Turbulence Observations in the Stratosphere (LITOS), which captured, for the first time, the entire turbulence spectrum down to the viscous subrange in the stratosphere [30,31]. HYFLITS has also developed and flown a payload with hot-wire anemometry and cold-wire thermometry. However, the payload was designed to be inexpensive and one-time use, where measurements are taken only during the descent phase for the purpose of minimizing the effects of the balloon wake, and real-time data telemetry was achieved at the cost of sample rate [32]. FiSH’s design is closer to the LITOS payload which also uses very high sampling rates, and all data are stored on the payload for post-flight spectral analysis, thus requiring payload recovery. One unique feature of FiSH is its ability to measure macroscopic and turbulent parameters in high temporal and spatial resolution for significant flight range, altitude, and duration.
2.1.2. Generalized Aerosol Sampling Payload (GASP)
We attempt here, for the first time to our knowledge, to collect filter paper samples of aerosols directly from the mid-stratosphere (defined here as altitudes above 23 km). GASP is an aerosol sampler designed to collect multiple filter paper samples at programmable altitudes during balloon ascent/descent and at a targeted “float” altitude above 23 km. The sampler was developed for BIRDIES by Lawrence Livermore National Laboratory (LLNL) and UMD with the bulk of the fabrication and testing performed by UMD. The payload pumps a large volume of air (~1 m3) through a series of filter paper samples, collecting all particles of size range > 0.015 µm on the filters. The payload is retrieved after flight and the filter paper samples are sealed and sent to LLNL for beryllium-10 and -7 analysis at their AMS facility called CAMS (Center for Accelerator Mass Spectrometry). Once opened at CAMS, the contents are immediately acidified in the presence of a beryllium carrier that consisted of isotopically pure 9Be. Note that this methodology does not introduce any chemical changes or loss during sample preparation or laboratory analysis.
The 5 kg weight payload consists of a custom-designed box that is sealed before flight and has four exterior holes for sample collection. Behind each hole is a 3D-printed filter housing, each of which contains a sterile fine-mesh filter paper. Behind each paper sample is a rubber tube that connects to an electronic actuator. Inside the box are the pump, onboard GPS unit, internal temperature sensor, power supply, and stack of Arduino chips that work as a flight computer which records ambient conditions (pressure, temperature, and altitude) and controls the operation of the pump (Figure 3a,b). The channel leading up to the pump contains a differential pressure sensor to monitor the flow rate of the sampled air. The sampled air is then exhausted out of the payload. The payload is designed to take an air sample on the ground through channel 1 when it is first powered up. This 10 min observation serves as a ground reference point. Once the sample is taken, the filter on channel 1 is manually replaced with a new filter and the payload is sealed for flight. The remaining samples are triggered by altitude via the onboard GPS unit. After the payload is launched, a sample is collected through channel 1 once the balloon rises above a programmable altitude, most often set to 20 km or higher, i.e., the “float altitude”. This sample is collected over a 20 min period. The tandem-balloon configuration allows the payload to spend approximately two hours in the mid-stratosphere and the last three samples are collected each with a 20 min collection duration and 30 s pause between samples.
The sample periods were chosen based on a desired air volume. Early testing suggested that we would sample 1.5 kg of air in ten minutes on the ground and 1 kg of air in twenty minutes at altitude. One of the challenges we faced with designing GASP was to have it function in an atmosphere that was roughly 1% of surface pressure. A low-pressure environment affects the efficiency of the pump because air flow is driven by pressure differentials. The pressure inside the pump needs to be lower than the ambient pressure to pull air in. In addition, when the air density is low, cooling (convection) of the pump becomes more difficult and the pump is prone to overheating. We estimated the required air volume sampling efficiency for collecting quantifiable amounts of beryllium-10 in the mid-stratosphere based on 10Be samples collected during prior aircraft flights in the lower stratosphere [18], which suggested that ~106 atoms/m3 of both 10Be and 7Be should be present at these altitudes. In addition, we analyzed stored aerosol samples taken on four flights over California and obtained from NASA Ames. These samples were taken at ~12 km cruise altitude from a DC-8 aircraft in 2017 and processed at LLNL CAMS for beryllium-10 in 2021. The samples showed 10Be concentrations at 12 km ranging from 1.9 to 4.8 (105 atoms/kg air). Ascent and descent altitudes (0.3 to 11 km), in contrast, ranged from 0.7 to 1.8 (105 atoms/kg air). The total air mass sampled by each filter ranged from 0.4 to 0.6 kg, and 10Be analyses were successful using only ~10% subaliquots of each collection filter. These measurements supported the decision to target 1 kg of air per GASP sample since 105–106 atoms 10Be (or 7Be) is sufficiently above measurement backgrounds at CAMS for these isotopes [12].
2.2. Flight Operations
BIRDIES flight missions were a collaboration between LLNL and the students and researchers at UMD. A unique feature of BIRDIES was the inclusion of five undergraduate students from the UMD Balloon Payload Program (
The tandem-balloon configuration was designed to achieve a cruise or float flight. This configuration was motivated initially for measuring 3 d fluctuations with 3 d hot-wire sensors along horizontal disturbances but is also useful for aerosol sampling when the interest is on a longitude–latitude scan, rather than just an ascent and descent (i.e., altitude) scan. Our tandem-balloon configuration consists of two Kaymont 1600 g balloons, each connected to the flight train with a separate flight termination device on a 10 m line (Figure 3c). One balloon acts as the ascent balloon and is terminated once the payload has reached the desired float altitude, e.g., 22 km. The second balloon also helps with the ascending phase, but additionally acts as the float balloon and is designed to be neutrally buoyant, although the accuracy of this is dependent on the error between our pre-flight calculations and the ambient environment aloft. The tandem-balloon configuration also includes a larger parachute (i.e., 10 ft diameter), a command module, and a radar reflector, and is often flown with an IMET-4 radiosonde to gather data about the air pressure, temperature, and moisture environment. GASP is located at the end of the flight train on a 10 m flight line below the parachute. Once the balloon platform is airborne, no changes can be made to the amount of helium in the balloons or to the total flight train weight to adjust the altitude. The only change that can be made is to terminate the flight by activating the flight termination devices.
Ideally, GASP and FiSH would be flown on the same balloon platform such that the turbulence measurements are co-located with the aerosol collection, telling us something both about the intensity of turbulence and the possible dynamics that are producing it, e.g., using beryllium as a tracer for detecting gravity waves. This concept is similar to the HYFLITS payload which integrates an optical particle counter (for characterization of the aerosol size environment) with the hot- and cold-wire anemometry into a single, small light-weight payload. In contrast, GASP is a large payload which would create a substantial wake if flown alongside FiSH. In order to obtain high-quality turbulence measurements, FiSH requires an undisturbed environment. The fact that the payloads are not co-located on the same balloon platform is somewhat mitigated by launching the two payloads quickly in series. The time lag between launches was on the order of an hour and was limited by the time it took to inflate the balloon and ready the flight train.
A schematic of our pre-flight, flight, and post-flight procedures are shown in Figure 4. More details are covered in [28]. Preparation activities included the selection of launch day, the determination of desired flight characteristics, defining personnel responsibilities, payload preparation, and filing Federal Aviation Administration (FAA) notifications and approvals. Planned flight characteristics included the determination of a desired ascent rate, apogee, flight path, and neck lift. The desired peak altitude for the single-balloon configuration was usually around 20–30 km but depended on ambient atmospheric conditions and the predicted flight path and estimated flight train weight. Launch day was selected based largely on weather (e.g., we needed low to moderate surface wind and gust speeds and a favorable wind direction for launch as well as no significant precipitation), but also based on the chance that we would collect measurements in the aftermath or immediate wake of a gravity wave (e.g., by targeting the time after major deep convection storm events). Other criteria included flight train readiness and that we did not interfere with DCOTSS’ ER–2 flight plans since we shared hanger space. Minor damage to the payload and flight train was not uncommon during BIRDIES landings, and some time was also needed for repairs between flights.
2.3. Precipitation Collection
Given that 7Be and 10Be are removed from the atmosphere primarily by wet deposition [33,34], a precipitation collection campaign was added to BIRDIES as a relatively easy way to obtain measurements of 10Be/7Be ratios across a spatial domain. We used passive rainfall collectors designed and built by LLNL. These units, called Raincubes, were designed to collect up to 20 L of precipitation using a collection area of 91 cm × 91 cm. The Raincubes were designed such that 25 mm of precipitation would produce a 19-L sample. A Raincube consists of twelve pieces of polyvinyl chloride (PVC) pipe, each 76 cm in length, connected by eight PVC side outlet elbow fittings to form a cube. Precipitation is funneled into a sterile 20 L HDPE carboy (Figure 5a).
The samples were manually collected after six wet weather system events within a three-week span to examine temporal, or event-scale, variability. The units were readied for the next storm with a new sterile carboy. To mitigate the potential for terrestrial source (e.g., dust) contamination of 10Be in our precipitation samples, we flushed each Raincube’s funnel and tubing between sampling periods using four liters of distilled water and elevated the collectors off the ground. One Raincube was deployed at each of eleven ARM SGP sites in northern Oklahoma (Figure 5b) and one at the airport in Salina, Kansas. The Kansas site was co-located with the BIRDIES balloon launch location and offered a comparison point that was farther away from the ARM domain, but still within the same climatological region. Our eleven ARM collection sites were distributed over an area of ~10,000 km2 to capture spatial variability on the scale expected for convective storm formation. All locations had one or more rain gauges operated by the DOE ARM program [35] and our naming convention follows the ARM site names. The location in Salina (SAL) (240 km north of the DOE ARM Central Facility) had an airport-operated rain gauge. The exact locations, collection times, and measured precipitation amounts for each site are listed in [36,37].
Because BIRDIES was a relatively short campaign, we also evaluated the continuous measurements of airborne 7Be from the nearest International Monitoring System (IMS) radionuclide station (RN74) to obtain a reference for daily variability. IMS uses gamma spectrometric analysis of daily particulate air monitoring to detect 7Be activity concentrations. Daily 7Be measurements are available from RN74 in Ashland, Kansas (37.2° N, 99.8° W), which is ~200 km west–northwest of ARM SGP and ~250 km southwest of Salina Airport (Figure 5c). These data are presented to illustrate a longer record of temporal variability in the region, to provide continuous daily observations of 7Be during our BIRDIES campaign, and to compare to GEOS-Chem-simulated values of 7Be to test model performance.
2.4. Laboratory Analysis
Accelerated Mass Spectrometry (AMS) is needed for long-lived isotopes 10Be and 36Cl, but 7Be is typically measured via gamma counting. For BIRDIES, 7Be/9Be and 10Be/9Be were measured on the same AMS target material that contained a known amount of stable 9Be. This approach simplified quantification of 10Be/7Be and each sample analysis was completed within 10–20 min, enabling large, high precision beryllium isotope datasets. Figure 6 illustrates the sample preparation and analysis steps for both the precipitation and GASP samples. Other species analyzed in the precipitation samples included cosmogenic isotopes (3H, 36Cl), stable isotopes (δ18O, δ17O, δ2H), and major cation and anion concentrations (Na, Mg, Al, K, Ca, Ti, Mn, and Fe). Noble Gas Mass Spectrometry (NGMS), which measures 3He accumulation from the decay of 3H, was used to obtain low-level 3H (tritium) measurements [38]. The stable hydrogen and oxygen isotopes were analyzed with a Picarro L2140-i isotopic water analyzer.
The large suite of species was analyzed to provide additional context for the wet deposition Be measurements and to help ascertain potential terrestrial contamination. 36Cl and 3H are atmospherically produced cosmogenic isotopes that are water-soluble but less particle-reactive than beryllium. As such, they are removed from the atmosphere via wet deposition like beryllium, but do not conservatively track aerosols. 36Cl and 3H also do not adhere to and concentrate in surfacial materials like beryllium and are therefore less likely to be remobilized from previous precipitation events and accompany terrestrial-based particles. Positive correlations between 10Be, which can accumulate in near-surface materials for thousands of years, and species such as Mg, Ca, and Al, on the other hand, would suggest the presence of a terrestrial source (i.e., contamination) for beryllium in our rainfall samples, resulting in a 10Be/7Be bias toward higher ratios.
2.5. Models
2.5.1. GEOS-Chem
GEOS-Chem, a global 3 d chemical transport model driven by Modern Era Retrospective analysis for Research and Applications version 2 (MERRA2) meteorological reanalysis data, was run to simulate daily 7Be and 10Be concentrations on our BIRDIES domain. The model was run at a spatial resolution of 2 degree × 2.5 degree, and at 72 levels for vertical resolution. For transport/convection, GEOS-Chem was run at 600 s temporal resolution, and for chemistry/emission, at a 1200 s resolution. Instead of using the default beryllium production rate by [7], we ran GEOS-Chem with updated production rates for beryllium provided by [39]. These updated production rates are 3-dimensional and temporally varying, and are based on the latest production model called CRAC:BE (Cosmic Ray-Induced Atmospheric Cascade for Beryllium) [14]. The new production rates use the solar modulation data by [40] and are based on a realistic estimate of the Earth’s geomagnetic field shielding effect.
2.5.2. Agglomerative Clustering
Agglomerative clustering is a machine learning technique in which each sample in a dataset is combined with another sample or group of samples, by similarity, iteratively, until the entire dataset is organized into a hierarchy. Agglomerative clustering with the median linkage function was performed on the precipitation dataset. Metrics included the calculation of the Pearson correlation coefficient (R) for each dimension (analyzed concentration of isotope or chemical species) of the dataset. The total dataset consisted of 64 collected precipitation samples, with each sample containing 23 analytical measurements (or dimensions). Out of the 64 samples, one was a sample of the distilled water used to flush the Raincubes and considered the test blank, FB-1, and is used as the control sample that the others are compared to. Eight out of the sixty-four samples did not have values for all dimensions, so they were dropped and not included in the agglomerative clustering. This left 56 23-dimensional samples, including FB-1.
3. Results and Discussion
3.1. Overview of Campaign
Balloon flight operations were executed on eight days, starting on 27 May and ending on 12 June 2022. This time period overlapped with the precipitation collection period (21 May to 9 June 2022), as well as a few relevant DCOTSS ER–2 flights, including a science flight over Oklahoma on the night of 31 May–1 June. Targeted weather events during BIRDIES are listed in Table 1 along with the timing of the storm(s), total collection duration, and average amount of precipitation that fell at the Raincube sites during each collection period (i.e., Event). Ideally, each collection period would have only included one storm system; however, our manual collection strategy was not optimized for this and all but Event 4 included more than one system. As an example, Figure 7 illustrates this for three of the ARM collection sites, which show the timing of the storm systems that produced precipitation across the domain. If an event had separate storms, they are designated in Table 1 as system 1 and system 2. Five of six of the collection events (Events 2–6) included evidence of overshooting convection. These five collections follow each other rapidly over a period of only eleven days. Overshooting activity is based on the echo tops reported by DCOTSS (when available, e.g., [41]) which uses gridded NEXRAD WSR-88D Radar data [42] or is based on the maximum GOES-derived cloud top height product calculated by the DOE ARM program. Note that because Event 1 contained more than two storm systems over a relatively long time (which were impossible to separate out once precipitation was collected), it is mostly dropped from further analysis.
The balloon flights, notated BIRDIES-03 through BIRDIES-10, and their scientific objectives and comments about the how the flights went are listed in Table 2. During BIRDIES, the turbulence payload FiSH was launched three times and had an average ascent rate of 5.6 m/s; GASP was launched a total of six times and had an average “float” ascent rate (also referred to as phase 2 ascent rate) of 1.5 m/s. GASP and FiSH were launched in series once during BIRDIES-05 and are listed as BIRDIES-05a and BIRDIES-05b. BIRDIES-01 was a test flight conducted in Maryland and BIRDIES-02 was aborted due to high winds.
BIRDIES payloads were flown during non-precipitation periods because of the weather constraints for launch; however, some of the flights were launched within 12 h after deep convective storm events in the region. Flight data are shown for two of the more successful flights, BIRDIES-05 and BIRDIES-07, in Figure 8. BIRDIES-05 GPS data from the 2 June flight indicate that the FiSH and GASPs were in good trajectory agreement (Figure 8a) until the float balloon on the tandem flight (BIRDIES-05a) failed prematurely at an altitude of 8 km (Figure 8b). After which, the remaining ascent balloon carrying GASP had a slow ascent rate (1.38 m/s) until descent starting at 18 km. Meanwhile, FiSH on BIRDIES-05b reached our target altitude of 26 km, and payload measurements were taken up to this altitude for a total of 77 min in the ascent phase before descent.
BIRDIES-07 was flown in between back-to-back deep convective storm events (Events 4 and 5, Table 2). The float balloon and GASP reached 26 km, following a slow ascent rate (labeled ascent phase 2 in Figure 8a) starting around 23 km, and “float” lasted 46 min (Figure 8b). While we approached neutral buoyancy with this float balloon, the error in our calculations was still significant, and the phase 2 ascent rate was 1.48 m/s.
3.2. High-Altitude Payload Measurements
BIRDIES-05b was launched about 24–36 h after significant storm activity in the region during the night of 31 May–1 June and on the morning of 1 June. FiSH measurements from this flight are presented in Figure 9 and Figure 10. Figure 9 shows the high-frequency hot and cold-wire anemometry and microphone acoustic pressure measurements. The power spectra using Welch’s method (also called the periodogram method) highlight measurements of 5 s time bins (~28 m length scale) at three altitudes, spanning the lower-to-mid-stratosphere (22–25 km). The wavenumber k, frequency f, and length scale L are related through , where w is the balloon ascent speed. High-resolution hot- and cold-wire measurements identified turbulent length scales in the inertial subrange (down to ~0.1 m with a −5/3 slope on the power spectrum with wavenumber as abscissa), but failed to observe the viscous subrange (a −7 slope on the power spectrum) due to noise (Figure 9a,b). The elevated noise seen above 100 Hz may be caused by the wake from the balloon flight train above the payload, from motions of the payload caused by non-ideal trim conditions, or from sensitivity of the constant voltage anemometer (CVA). An explanation for the latter is explored further in Section 5. High-resolution microphone measurements also captured turbulent length scales of acoustic pressure fluctuations in the infra-sonic and inertial subrange down to 1 m (Figure 9c). The infra-sonic range is indicated by a −1 slope whereas the inertial subrange is represented by a −7/3 slope. Here, the wavenumber is related to frequency by , where a is local speed of sound. While we observed an increasing slope at smaller length scales, no references clearly define the viscous subrange in acoustic pressure signals.
Low-resolution measurements from the pressure transducer and thermocouples from BIRDIES-05b are plotted in Figure 10. These include the Brunt–Väisälä frequency NB (s−2), wind shear S (s−2), and the Richardson number Ri. The potential temperature and wind shear are smoothed with a moving average over 25 m after differentiation. A smaller Ri indicates a lower stability of the flow, which indicates a higher likelihood of turbulence. We use the historical Ricrit = 0.25 to indicate times when the flow is likely turbulent (i.e., Ri < Ricrit). This showed that in our low-resolution observations, turbulence regions were observed intermittently throughout most of the altitudes flown. These measurements were taken the day after widespread severe overnight thunderstorms in Northern Oklahoma. While there is evidence of gravity waves being produced in the region from these storms [41], we cannot definitively say that our measurements captured the decayed overshoot turbulence given the time offset. Moreover, if we compare our range of Ri values to those taken by LITOS during a different gravity wave event [43], they report overall larger Ri numbers than our observations and fewer incidences of Ri < Ricrit. While our data were taken in a different location and time, our methodology may introduce a bias in our Ri numbers towards smaller values. BIRDIES Ri numbers are based on a length scale of around 25 m in altitude (much smaller than LITOS), and it has been noted that computing Ri on a smaller scale yields locally smaller Ri numbers [43].
Analytical results for 10Be and 7Be measured on n = 26 GASP filter samples are shown in Table 3. Few samples returned measurable 10Be (n = 7) or 7Be (n = 2) above laboratory backgrounds (1.4 × 104 atoms for 10Be and 1.5 × 103 atoms 7Be, as determined from full-chemistry process blanks), which were significant to all beryllium isotope measurements and generally of the same order of magnitude. Samples with the highest measured concentrations were less than a factor of four above their respective blanks, resulting in large 1σ uncertainties for calculated concentrations (between 10 and 85%). Notably, these measured concentrations were ~100× lower than our collection target (~106 atoms for either isotope). Only the ground sample acquired during BIRDIES-09 returned measurements for both isotopes, returning a measured ratio of 10Be/7Be = 2.9 ± 2.0. This value is on the high end when compared to the ratios returned from the rainfall collection campaign (the closest event timing was Event 6), although this GASP ground value does have a large amount of uncertainty. Aside from the BIRDIES-09 ground sample, all detectable measurements for 10Be were from samples collected at altitudes in the stratosphere (15–25.5 km) and were exclusive to the BIRDIES-07 and -09 campaigns. 7Be was detected in only one sample at altitude, which was collected from the upper troposphere (10–11.7 km) and during the abbreviated ascent of BIRDIES-04.
We hypothesize two primary reasons why most of our GASP samples showed negligible to immeasureable amounts of 7Be and 10Be. First, the inflow rate was likely too slow over the sample duration, leading to insufficient aerosol collections. Second, GASP suffered from a pump overheating issue, which terminated the programmed sampling procedure mid-flight, which resulted in fewer samples being collected than expected. Recommendations for future designs are given in Section 5.
3.3. Ground-Based Isotope Measurements
Our wet deposition collection produced 64 individual precipitation samples brought back to CAMS for analysis. These varied in volume from 25 mL to 3.8 L. The upper limit size was restricted to 3.8 L (or 1 U.S. gallon) to save space during transportation back to LLNL, although collected samples were as large as 5 gallons (i.e., the volume of the sample container). If a collected sample was greater than 1 gallon, it was homogenized (i.e., well shaken) after collection and then reduced to a 1-gallon sample before transport. Analyzed beryllium measurements and 10Be/7Be ratios for all of the individual samples are reported in [36]. The campaign data are also downloadable from the DOE ARM campaign website [37].
Laboratory measured concentrations (+/− one standard deviation) of 10Be and 7Be appear to be positively correlated in our samples, i.e., a higher concentration of 10Be is usually associated with a higher concentration of 7Be (Figure 11). Overall, Event 2 has the lowest mean concentrations of both 10Be and 7Be and Events 4 and 5 had the highest concentrations. Event 3 has the smallest amount of site-to-site or spatial variability. Event 10Be/7Be ratios range between 1 and 2. The ranges include spatial variability as each event contains roughly 11 site samples across a domain of 10,000 km2. Our reported 10Be/7Be ratios fall within published values for altitudes reaching the troposphere and lower stratosphere, although they are all higher than the 0.6 ratio predicted for purely tropospheric air. However, this 10Be/7Be threshold may be too low even for surface air. Using GEOS-Chem, [39] suggests 10Be/7Be ratios of 0.9–1.0 are often found at the surface and can be as high as 1.4 during strong spring-time STE events in the northern hemisphere. Recent 10Be/7Be measurements (~1.5 to 2.5) in rainfall by [44] support this model prediction. Reference [44] connects these relatively high precipitation ratios to stratospheric intrusions. Our mean 10Be/7Be ratios range from 1.5 (Event 2) to 1.8 (Event 3), which also likely show some influence of stratospheric air. Also note that our one GASP sample, which had measurable [10Be] and [7Be] from the ground surface, showed a 10Be/7Be ratio of 2.9 ± 2.0, which is also well above the 0.6 value.
In Figure 11, the measured concentrations on the second y-axis are in Be atoms/g water while the GEOS-Chem model output is in mol/m3 air. These are proportional but not easily translatable units. Thus, for comparing expected and observed concentrations, the pattern is useful while the absolute magnitudes are not. Predicted GEOS-Chem daily 10Be and 7Be concentrations and ratios show fairly high trend agreement with the event observations. A few exceptions are the modeled 10Be concentration for Event 2 (it is overpredicted) and the underpredicted 10Be/7Be ratio for Event 3 (due to the overprediction of 7Be concentration). Note that GEOS-Chem is not able to resolve spatial variability given its coarser grid size and that only one daily value is predicted for the entire BIRDIES domain.
The GEOS-Chem-predicted temporal variability in [7Be] and [10Be] follows the event measurement trends with fairly good agreement, suggesting that our event-scale variability may be controlled by regional- or synoptic-scale variability, and, to a lesser degree, by individual storm cloud overshooting events which were not resolved by GEOS-Chem. It has been previously shown that isotopic composition in precipitation is influenced by synoptic weather patterns [45]. These include atmospheric conditions at the moisture source, moisture transport trajectories, mixing between vapors from different origins, and subcloud processes (e.g., re-evaporation and convection).
To assess the extent of and causes of our beryllium variability, we first look at temporal variability by comparing our measured BIRDIES [7Be] measurements with a longer data record collected at the IMS radionuclide station in Kansas. These data periods are also simulated with GEOS-Chem, centered over the IMS Kansas site. Figure 12 shows the daily time series for observed Kansas [7Be] data for 2021 through mid-summer 2022, and the differences between the modeled values and observations. Overall, we see best agreement between the observed and modeled values in the boreal winter and spring months. With some exceptions, the highest disagreement is found in the summer months (June through August), when the model appears to overestimate [7Be]. This may be due to the model overpredicting the amount of 7Be transported from the upper troposphere to the ground via midlatitude convective circulations, which are most active in the summer months [23,46,47]. On the BIRDIES experiment days, the model also overpredicted daily [7Be] compared to both the precipitation (Figure 11) and IMS (Figure 12) measurements.
To look more closely at spatial beryllium variability, we focus on Event 2. Event 2 overlapped in time and space with the DCOTSS 31 May–1 June ER–2 aircraft mission that focused on tracking night-time deep convection storms over northern Oklahoma and provided additional high-resolution measurements across our domain. This storm event was also followed by our morning BIRDIES-05 flight on 2 June, as discussed above. Overshooting convection activity in the area starting around 22:00 UTC (17:00 local time) on 31 May in western Oklahoma and ending ~06:30 UTC (01:30 local time) on 1 June when the storm had moved eastward. Widespread overshooting in the region reached a maximum echo-top height of 19 km, and sustained echo tops near 17–18 km for the majority of the night-time storm event [41]. During this storm, we measured a large range of 10Be/7Be values in our precipitation samples (1.1 to 2.1) with the highest value (an outlier) measured at E32, one of our northern sites (Figure 13). Radar identified a severe storm over E32 around 00 UTC. The majority of the sites in our domain, on the other hand, received precipitation in smaller amounts later on 1 June, during the second storm system (see Figure 7).
Our interpretation of Be in the precipitation samples assumes that the isotopes are primarily coming from atmospheric scavenging, i.e., wet deposition, and that the measured beryllium isotope composition should reflect the vertical column of aerosol particles in the overlying atmosphere. This assumption that Be was not contaminated by ground sources was tested by analyzing the samples for terrestrial species Mg, Ca, Al, K, Ti, Mn, and Fe, which would indicate terrestrial contamination (a concern for long-lived 10Be). The Pearson correlation coefficients (R-value) between 10Be/7Be and each of these terrestrial species ranged from 0.01 to 0.13, indicating weak correlations (Figure 13). This suggests that the beryllium isotopes in our precipitation samples were not contaminated by terrestrial Be sources (e.g., from soil or vegetation) and instead are likely cosmogenic in origin. This is also confirmed by the positive correlations found between [7Be] and [10Be] and [36Cl] (R= 0.34, R= 0.41, respectively), since all are cosmogenic in origin. This positive correlation has been found before in precipitation samples, although [24], for example, report even higher cross-correlation values of r = 0.7 ([7Be] and [36Cl]) and r = 0.8 ([10Be] and [36Cl]). We also found a positive correlation between 10Be/7Be and cosmogenic 3H (R = 0.55), and to a lesser extent, we found negative correlations between 10Be/7Be and δ18O (−0.24), δ17O (R = −0.25), and δ2H (R = −0.24).
Prior research has shown that the concentration of tritium in precipitation is positively correlated with the fraction of convective precipitation [48]. In [48], the authors show that condensation and riming associated with atmospheric moisture produces higher tritium concentrations when the riming occurs in deep convection with entrained air at higher altitudes. This is consistent with our observations that high tritium in the BIRDIES samples is correlated with higher concentrations of beryllium isotopes, and larger 10Be/7Be ratios are associated with stratospheric sources of beryllium. However, [48] also found that convective precipitation is associated with higher [δ2H] and [δ18O]. Since we found weak, negative correlations between 10Be/7Be and these species, we hypothesize that our negative correlations found during BIRDIES could be the result of distinct sources of vapor contributing to stratiform and convective precipitation events, as discussed next.
Agglomerative clustering using a median linkage function was performed on the precipitation dataset until the entire dataset was organized into a hierarchy. The results are visualized in the matrix in Figure 14. Choosing a different linkage function other than median will change how the samples are clustered, but similar patterns were seen when trying other linkage functions. In the matrix, the columns are the dimensions (i.e., analyzed chemical species or sample properties), the rows represent individual precipitation samples, and the colors represent the z-score for the entry, normalized by column. Z-scores represent standard deviations, e.g., a high z-score at a given row and column means that the value is far away from the column mean. The dendrogram on the left side of the figure shows the hierarchical clustering of each of the samples. Note that only select species were included in the clustering shown here, and they include the following: 10Be, 7Be, 10Be/7Be, 36Cl, Cl, 3H, 10Be/36Cl, 7Be/36Cl, δ18O, δ2H, Na, and Al.
The clustering shows that the precipitation samples are broadly classified into three clusters or groups which collectively contain nearly 95% of the samples. The remaining 5% include the field blank (FB-1), E36-1b, E33-2, and E36-5. Note that E33-2 and E36-5 were very small sample sizes (i.e., very low precipitation amounts) and E36-1b was a second collection during that storm event. The main three clusters for all other samples are labeled A, B, and C (Figure 14). The clusters are largely arranged by Event number and not by geographic location. In fact, the Salina, Kansas samples are more similar to the adjacent storms in northern Oklahoma than they are to each other. Also, the Oklahoma sites have no apparent geographic clustering even though they are spread over a domain of 10,000 km2, which includes a climatological west-to-east precipitation gradient. Cluster A and Cluster B both contain samples from Events 1, 3, 4, and 6. Cluster C contains Events 2 and 5 and one sample each from Event 4 (E11-4, a very small sample size) and 6 (I9-6). Given this, Events 2 and 5 are most closely related in species composition and Events 1, 3, 4, and 6 are most closely related.
The z-scores somewhat help explain the species trends that we saw during the experiment. Convective precipitation associated with air masses originating at higher latitudes or higher altitudes (e.g., Rocky Mountains) would bring different δ2H and δ18O values in precipitation than stratiform precipitation associated with air masses originating in the Gulf of Mexico [48]. These different moisture sources also partially explain variability in tritium, with lower tritium concentrations associated with stratiform precipitation from the Gulf of Mexico and higher tritium concentrations associated with high latitude or high-altitude moisture sources [49]. We ran backtrajectory ensembles using NOAA’s Hysplit model with High-Resolution Rapid Refresh (HRRR) 3 km meteorological data [50] for each of our event periods to assess any differences in moisture transport and origin that may explain our event-scale variability. These results are summarized in Table 4. Air mass origin is based on the majority agreement from the trajectory ensembles, which were run from starting altitudes ranging from 100 m to 5 km. These trajectory ensembles show primarily either a Gulf of Mexico or Pacific Ocean moisture source for our samples (Table 4).
Gulf of Mexico origin ensemble backtrajectories were found to have good agreement with each other during our Event 4, and to a lesser degree during Events 2, 5, and 6 (Table 4). Event 3 was Pacific Ocean in origin. As Event 4 saw lower-than-median 3H values, and higher-than-median δ2H and δ18O values, this finding appears to corroborate an air mass originating from the Gulf of Mexico. Much-lower-than-median 3H values were found in Event 2, also suggesting a Gulf of Mexico origin. But since Event 2 and Event 5 were statistically most similar in species composition but may have had different moisture origins, moisture origin does not appear to tell the entire story. Elucidating the exact mechanisms that produced our temporal variability has proven to be difficult, and at this point, we cannot definitively say that our 10Be/7Be ratios in precipitation are linked to STE processes resulting from deep convection or are due to different moisture sources. Instead, our measurements suggest a combination of the two factors for explaining our isotopic compositions.
4. Conclusions
BIRDIES was motivated by the lack of direct measurements of 10Be and 7Be in the mid-stratosphere and the uncertainties surrounding 10Be/7Be as a function of altitude, and how it varies in space and time over short time scales (i.e., storm systems) due to STE dynamics. BIRDIES flew eight high-altitude balloon missions aimed at reaching mid-stratospheric heights with two custom-made sensors for measuring turbulence and for obtaining aerosol samples. Although these balloon missions had a number of failures and neutral buoyancy in the mid-stratosphere was not fully achieved with the tandem-balloon design, we were able to obtain observations of the Brunt–Väisälä frequency, wind shear, and the Richardson number during flight ascent up to altitudes of 28 km with the FiSH payload. Our observations show elevated turbulence in the mid-stratosphere following a series of convective storms in the area. Linking this directly to evidence of a gravity wave event, however, was not clear since the measurements were taken about 24–36 h after the overshooting storm events in the region. GASP proved more problematic at these high altitudes and our retrieved samples contained too few aerosols for beryllium analysis. Ground-based beryllium collection was successfully performed with a precipitation campaign that utilized the DOE ARM SGP facility. These samples were taken across five precipitation events with observed cloud overshooting and across a large domain (10,000 km2) and within a short number of days (n = 11) that would eliminate any kind of seasonal variability. We took advantage of laboratory capabilities at LLNL and also ran these samples for terrestrial species and other cosmogenic isotopes. The results clearly showed that the beryllium in our precipitation samples was cosmogenic in origin and not contaminated by terrestrial sources. Correlations with beryllium and 3H and 36Cl, as well as with the hydrogen and oxygen isotopes of δ18O and δ2H, were less straightforward, and hint at the fact that our 10Be/7Be ratios likely resulted from a combination of unique moisture sources and STE dynamics. Localized deep convection was harder to pinpoint as the cause for our 10Be/7Be variability due to the fact that overall event-scale variability appeared to be greater than spatial-scale variability and the fact that GEOS-CHEM was able to predict our trends in event ratios with a high degree of accuracy. GEOS-CHEM is unable to resolve individual storm clouds. We did have one sample, E32-2, which may have shown evidence of deep convection given its high 10Be/7Be ratio (2.1) and the fact that there is evidence of deep convection and cloud overshooting very close to the site from the DCOTSS flight mission and radar observations. In conclusion, while BIRDIES did not produce robust 10Be/7Be measurements taken directly in the mid-stratosphere as we had aimed, the experiment did show that ground-based precipitation collection can be used to measure cosmogenic isotopes without terrestrial contamination and that these measurements show a high degree of temporal variability even over very short time scales, i.e., back-to-back storm events.
5. Lessons Learned and Recommended Future Work
During BIRDIES, we experienced problems with the payloads and flight operations, which led to recommendations on how to improve future high-altitude balloon missions. In this campaign, the tandem-balloon operations did not achieve the desired float altitude. This may have been caused by our assumptions used to calculate neck lift such as drag coefficient and uncertainties associated with neck lift measurement during inflation. Other times, it was simply due to physical failure either with the balloons or flight train swivels. Improvements with calculation error may be achieved with modified drag coefficient based on existing flight data, or simulations achieved with computational fluid dynamics tools. Alternatives to the tandem-balloon design could also be investigated. These include using a heliotrope (limited to a lighter payload), which is a solar-powered balloon designed for long-duration float flight [51], or a control device to vent off helium during flight according to real-time ascent rate data, such as described in [32].
While FiSH collected a large amount of macroscopic and turbulence data during our campaign, we later determined that the hot-wire and cold-wire CVA were not performing consistently. Whereas earlier attempts highlighted the success of capturing a viscous subrange of turbulent velocity and temperature in the stratosphere with FiSH [27], measurements taken during BIRDIES have more noise. This may be due to errors in the overheat ratio in the CVA system. We set the overheat ratio on the ground; however, this ratio actually changes with altitude during flight. As a result, a careful selection of the initial overheat ratio on the ground is needed to allow sufficient sensitivity of the sensor during flight while avoiding wire burnout at higher altitudes. This introduces uncertainties on setting up the overheat ratio as flight conditions vary from operation to operation (i.e., a lower overheat ratio on the ground leads to a lower overheat ratio at high altitudes). One suggestion is to switch to a constant temperature anemometer (CTA) for ease of operation, which possesses a constant operating wire temperature, but at the cost of sensitivity.
GASP, a novel payload designed for collecting first-ever measurements of 10Be and 7Be in the mid-stratosphere, would benefit from a co-located optical aerosol particle counter, e.g., POPS, to trigger sample collection, especially in an environment low in aerosols such as the stratosphere. Additionally, a more efficient pump is needed to pass more air through the filter paper samples. For example, we suggest sampling 1 kg of air over 30 min (at 18 km altitude this corresponds to approximately 0.004 m3/s). Flight costs and necessary support would be reduced by redesigning GASP to decrease its size and weight so that it can fly on alternative high-altitude balloons, such as the heliotrope described above. This would also increase measurement collection time in the stratosphere as heliotrope flights routinely last ten or more hours, depending on the amount of daylight hours. Other applications that would benefit from GASP’s improved capabilities include sampling forest fire plume chemistry and observing aerosol characteristics for the testing of climate geoengineering mitigation strategies.
Lastly, a future precipitation collection would benefit from automation if performed for an extended period. Our Raincube methodology was very labor-intensive, especially due to our goal of capturing back-to-back convection storms across a relatively large domain. Triggering the precipitation collection based on rain gauge data would lead to more accurate collection strategies.
Conceptualization, S.W., A.J.H., T.S.E., W.Z., S.N.S., P.J.C.-S. and H.B.; methodology, S.W., A.J.H., T.S.E., W.Z., S.N.S., P.J.C.-S., E.O., A.V. and M.R.; software, W.Z., S.N.S., M.G. and S.M.; validation, A.J.H., T.S.E., W.Z. and S.N.S.; formal analysis, S.W., A.J.H., T.S.E., W.Z., N.G. and J.M.L.; investigation, S.W., A.J.H., T.S.E., W.Z., E.O., A.V., S.M. and M.G.; resources, A.J.H., W.Z., S.N.S., E.O., M.R. and A.V.; data curation, S.W., A.J.H., W.Z., J.M.L., N.G., M.G. and S.M.; writing—original draft preparation, S.W., A.J.H., W.Z., J.M.L. and N.G.; writing—review and editing, S.W., A.J.H., W.Z., N.G. and J.M.L.; visualization, S.W., A.J.H., W.Z., N.G. and J.M.L.; supervision, S.W., A.J.H., T.S.E. and W.Z.; project administration, S.W. and A.J.H.; funding acquisition, S.W. and A.J.H. All authors have read and agreed to the published version of the manuscript.
Not applicable.
Not applicable.
BIRDIES 10Be and 7Be data are hosted by the DOE ARM user facility (
BIRDIES thanks the UMD BPP student team and BPP faculty mentor, Mary Boden, for their support and assistance for the balloon flight missions. Additionally, thanks go to Stuart Laurence for his support for the payload design and testing at UMD. An additional thank you goes to everyone who helped with the logistics for the balloon flights, including Michelle “Shelli” Swanson at the Salina Regional Airport and Mike Gonigan at LLNL for obtaining helium during a severe nation-wide helium shortage. The Atmospheric Radiation Measurement (ARM) SGP staff were critical to the success of the precipitation campaign and special thanks goes to John Schatz and Chris Martin. ARM, a U.S. Department of Energy (DOE) Office of Science user facility, is managed by the Biological and Environmental Research Program. Lastly, we would like to thank the DCOTSS researchers and planners, including Dan Chirica, Chad Homeyer, and Rei Ueyama, who gave us support at the Salina airport and served as collaborators. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
![View Image - Figure 1. Schematic of berylium-7 and -10 production and wet deposition as well as estimated ratios of 10Be/7Be in the stratosphere and troposphere. Our airborne (GASP) and ground-based wet deposition (Raincube) aerosol collection methods are also illustrated. Figure modified from [17].](https://media.proquest.com/media/hms/THUM/2/PTT2b?_a=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%3D%3D&_s=VYTS5OAV%2BzkZoVVSy4z6VdZ1Bvw%3D "View Image - Figure 1. Schematic of berylium-7 and -10 production and wet deposition as well as estimated ratios of 10Be/7Be in the stratosphere and troposphere. Our airborne (GASP) and ground-based wet deposition (Raincube) aerosol collection methods are also illustrated. Figure modified from [17].")
Enlarge this image.Figure 1. Schematic of berylium-7 and -10 production and wet deposition as well as estimated ratios of 10Be/7Be in the stratosphere and troposphere. Our airborne (GASP) and ground-based wet deposition (Raincube) aerosol collection methods are also illustrated. Figure modified from [17].
Enlarge this image.Figure 2. (a) External view of FiSH showing the cold- and hot-wire anemometers and tail fin; (b) internal view of FiSH showing the onboard electronics; (c) schematic of the single-balloon configuration with the FiSH payload; (d) photograph of FiSH launched during BIRDIES.
Enlarge this image.Figure 3. (a) GASP external and internal views; (b) close–up view of GASP’s interior showing the onboard electronics, pump, and filter system; (c) schematic of the tandem-balloon configuration with GASP; (d) photograph of GASP launched during BIRDIES.
Enlarge this image.Figure 4. Schematic of major pre-flight, flight, and post-flight activities for BIRDIES. These steps are for both the single- and tandem-balloon platforms.
Enlarge this image.Figure 5. (a) Photograph of the Raincube at E36 which shows a typical deployment during BIRDIES; (b) Raincube deployment map (blue stars) in the DOE ARM SGP domain centered on the Central Facility; (c) regional map showing central Oklahoma and Kansas and the locations of the Salina airport (SAL) and the ARM Central Facility (CF) Raincubes (as well as the boundaries of the ARM Raincube domain), and the IMS station RN74.
Enlarge this image.Figure 6. Schematic of the Accelerated Mass Spectrometry (AMS) analytical procedure for 10Be and 7Be including the sample preparation for the precipitation (Raincube) samples and aerosol (GASP filter) samples. Also shown are the split samples for anion, cation, and isotopes of H and O analysis.
Enlarge this image.Figure 7. One-hour precipitation amounts measured at three of the ARM sites (E32, I10, CF) to show that some events contained more than one storm system during the collection duration. The sample collection events are labeled and highlighted in gray. Collection durations differ between sites because the collection was performed manually. Event 1 is not shown.
Enlarge this image.Figure 8. Example of (a) flight trajectory and (b) altitude information, shown for BIRDIES-05a (GASP), BIRDIES-05b (FiSH payload), and BIRDIES-07 (GASP).
Enlarge this image.Figure 9. BIRDIES-05b (a) hot-wire, (b) cold-wire, and (c) microphone power spectrum from FiSH shown for three altitudes in the lower-mid-stratosphere.
Enlarge this image.Figure 10. FiSH measurements of (a) Brunt–Väisälä frequency, (b) wind shear, and (c) Richardson number on the ascent phase from BIRDIES-05b. Altitudes plotted range from 10 to 26 km (upper troposphere to mid-stratosphere). Dashed line is the critical Richardson number, Ricrit = 0.25.
Enlarge this image.Figure 11. Measured mean (+/− one standard deviation) and modeled concentrations of 10Be, 7Be, and their ratio (10Be/7Be) by event number. Each measured event contains up to 11 ARM SGP Raincube observations.
Enlarge this image.Figure 12. (a) Time series of daily 7Be measured at the IMS RN74 station in Ashland, Kansas; (b) difference between GEOS-Chem modeled and observed value; and (c) difference focused on the BIRDIES period with the collection Events 2-6 highlighted. Note in panel (c) that the collection durations of Events 3 and 4 overlap with the daily IMS record as indicated by bracketed lines. Average concurrent BIRDIES precipitation collection events are highlighted in (c). The dashed box in (b) represents the BIRDIES period shown in detail in panel (c).
Enlarge this image.Figure 13. Pearson correlation coefficients (R) for a subset of the analyzed chemical species in the BIRDIES precipitation dataset. Precipitation is event– and site-specific and is the value measured by the rain gauge. Precipitation weight is the amount of water in each sample received at the laboratory.
Enlarge this image.Figure 14. Agglomerative clustering with a median linkage function applied to the BIRDIES precipitation dataset. The colors represent the z-score, where the standard deviation from the median is calculated by column. The clustering shows three main groups (Clusters A–C).
Brief weather description of the six BIRDIES precipitation events. Precipitation dates include timing of first precipitation record in the domain and timing of the last precipitation record in the domain for the event period.
| Event | Precipitation Start Date, UTC | Precipitation End Date, UTC | Site-Averaged Total Precip. (mm); Site Range Total Precip. (mm) | Weather Description | Evidence of Overshooting and Max. Cloud Top Height |
|---|---|---|---|---|---|
| 1 | 21 May 2022 | 25 May 2022 | 79.8; | Multi-day high-rainfall event from multiple slow-moving large-scale storm systems, weak-to-moderate instability across domain. | None |
| 2 | 31 May 2022 | 1 June 2022 | 11.6; | Cold frontal passage and overnight thunderstorms on 31 May–1 June, isolated and severe. A second more widespread system came through later in the morning on 1 June. | Yes |
| 3 | 4 June 2022 | 5 June 2022 | 19.3; | Mesoscale convection system. Nocturnal convection induced by higher terrain to the west in Colorado. | None for system 1 (4 June) |
| 4 | 6 June 2022 | 6 June 2022 | 27.6; | Strong instability formed strong supercells; otherwise, widespread convection from a cold front passage, including severe thunderstorms in the domain overnight. | Yes: 15 km |
| 5 | 7 June 2022 | 8 June 2022 | 1.9; | Overnight severe isolated storms associated with a mesoscale convective system induced by higher terrain to the west in Colorado/New Mexico. | Yes |
| 6 | 9 June 2022 | 10 June 2022 | 7.7; | Strong instability, very isolated overnight thunderstorms, and some severe in the eastern domain. | Yes |
Details for the eight completed BIRDIES flights from Salina, Kansas. Also listed is the timing of the precipitation collection events.
| Launch | Date (UTC) | Payload | Balloon | Objectives | Comments on Flight |
|---|---|---|---|---|---|
| BIRDIES-03 | 27 May 2022 | FiSH | Single | Turbulence measurements up to 28 km | Spinning of FiSH was too high and problematic to obtain turbulence data. Payload reached 28.4 km. |
| BIRDIES-04 | 31 May 2022 | GASP | Tandem | Float after 23 km, four beryllium samples | Float balloon broke off prematurely at 8 km due to a broken swivel. Payload reached 14 km. Ascent rate with remaining balloon was 2.03 m/s. |
| Event 2 Precipitation Collection | |||||
| BIRDIES-05a | 2 June 2022 | GASP | Tandem | Float after 23 km, four beryllium samples | Reached 18 km. Float balloon failed at 8 km and ascent rate with remaining balloon was 1.38 m/s. |
| BIRDIES-05b | 2 June 2022 | FiSH | Single | Turbulence measurements up to 29 km | One unwinder jammed. Payload reached 26 km. |
| Event 3 Precipitation Collection | |||||
| BIRDIES-06 | 5 June 2022 | FiSH | Single | Turbulence measurements up to 29 km | Ascent balloon burst prematurely at 14.3 km. |
| Event 4 Precipitation Collection | |||||
| BIRDIES-07 | 7 June 2022 | GASP | Tandem | Float after 23 km, four beryllium samples | Reached 26 km. Actual ascent rate at “float” was 1.48 m/s. |
| Event 5 Precipitation Collection | |||||
| BIRDIES-08 | 9 June 2022 | GASP | Tandem | Float after 22 km, four beryllium samples | Float balloon was cut off prematurely due to an erroneous cutdown setting after reaching 22 km. |
| Event 6 Precipitation Collection | |||||
| BIRDIES-09 | 11 June 2022 | GASP | Tandem | Float after 20 km, four beryllium samples | One sample collection ends prematurely. Reached 27 km. Ascent rate at “float” was 1.30 m/s. |
| BIRDIES-10 | 12 June 2022 | GASP | Tandem | Float after 15 km, one cascading beryllium sample | GASP collided with hangar due to a wind gust and no collections were made. |
Measured concentrations and ratio of 10Be and 7Be for the collected GASP samples. All 7Be concentrations are decay-corrected to collection data (t = 0).
| Sample ID | Collection Altitude (km) | Collection Duration 2 (min) | Est. Air Mass 3 (kg) | CAMS ID | [10Be] (103 Atoms) | [7Be] (103 Atoms) 4 | 10Be/7Be | |||
|---|---|---|---|---|---|---|---|---|---|---|
| µ | ±1σ | µ | ±1σ | µ | ±1σ | |||||
| BIRDIES-04 | ||||||||||
| 4-00 Control | NA | NA | NA | BE51622 | ND | ND | ND | ND | ND | ND |
| 4-01 Ground | 0 | 20 | 8.296 | BE51623 | ND | ND | ND | ND | ND | ND |
| 4-02 Ascent | 10–11.6 | 14 | unknown | BE51624 | ND | ND | 6.1 | 4.2 | ND | ND |
| BIRDIES-05 | ||||||||||
| 5-00 Control | NA | NA | NA | BE51625 | ND | ND | ND | ND | ND | ND |
| 5-01 Ground | 0 | 20 | 1.960 | BE51626 | ND | ND | ND | ND | ND | ND |
| BIRDIES-07 | ||||||||||
| 7-00 Control | NA | NA | NA | BE51627 | ND | ND | ND | ND | ND | ND |
| 7-01 Ground | 0 | 20 | 2.110 | BE51628 | ND | ND | ND | ND | ND | ND |
| 7-02 Ascent | 15–22.5 | 20 | 0.830 | BE51629 | ND | ND | ND | ND | ND | ND |
| 7-03 Ascent/Float | 22.8–24.9 | 20 | unknown | BE51630 | 7.0 | 3.8 | ND | ND | ND | ND |
| 7-04 Float/Descent | 24.9–23.6 | 20 | unknown | BE51631 | 4.2 | 3.6 | ND | ND | ND | ND |
| 7-05 Descent | 22.6–6.4 | 20 | unknown | BE51632 | ND | ND | ND | ND | ND | ND |
| BIRDIES-08 | ||||||||||
| 8-00 Control | NA | NA | NA | BE51633 | ND | ND | ND | ND | ND | ND |
| 8-01 Ground | 0 | 20 | 2.207 | BE51634 | ND | ND | ND | ND | ND | ND |
| 8-02 Ascent | 15–15.5 | 1.4 | 0.101 | BE51635 | ND | ND | ND | ND | ND | ND |
| 8-03 Ascent/Descent | 15.7–20.9 | 20 | 0.800 | BE51636 | ND | ND | ND | ND | ND | ND |
| 8-04 Float/Descent | 20.2–5.9 | 20 | 2.662 | BE51637 | ND | ND | ND | ND | ND | ND |
| BIRDIES-09 | ||||||||||
| 9-00 Control | NA | NA | NA | BE51638 | ND | ND | ND | ND | ND | ND |
| 9-01 Ground | 0 | 20 | 2.142 | BE51639 | 15.8 | 4.8 | 5.4 | 3.3 | 2.9 | 2.0 |
| 9-02 Ascent/Float | 15–20.9 | 20 | 0.888 | BE51640 | 42.5 | 4.6 | ND | ND | ND | ND |
| 9-03 Float | 21–23.1 | 20 | 0.222 | BE51641 | 6.8 | 3.0 | ND | ND | ND | ND |
| 9-04 Float | 23.1–24 | 10.8 | 0.003 | BE51642 | 19.0 | 3.7 | ND | ND | ND | ND |
| 9-05 Float | 24.1–25.5 | 20 | 0.003 | BE51643 | 13.6 | 4.4 | ND | ND | ND | ND |
| BIRDIES-10 | ||||||||||
| 10-00 Control | NA | NA | NA | BE51644 | ND | ND | ND | ND | ND | ND |
| 10-01a Ground 1 | 0 | 20 | data lost | BE51645 | ND | ND | ND | ND | ND | ND |
| 10-01b Ground 1 | 0 | 20 | data lost | BE51646 | ND | ND | ND | ND | ND | ND |
| 10-01c Ground 1 | 0 | 20 | data lost | BE51647 | ND | ND | ND | ND | ND | ND |
Note: 10Be and 7Be concentrations are background-corrected using full laboratory process blanks; ND = no detection, and implies analytical result was indistinguishable from process blanks. Uncertainties from AMS measurement, blank correction, and beryllium carrier concentration are included in each concentration uncertainty. All uncertainties shown are 1σ. (1) Ground sample BIRDIES-10 contained a cascaded filter assembly with decreasing mesh size: a = 12 µm, b = 1 µm, c = 0.015 µm. All other samples were collected on 0.015 µm filters. (2) Duration reflects the length of time the filter was exposed through an open channel. (3) Air mass estimate from differential pressure sensor. Quantity is “unknown” when pump fails for some portion of collection. Damage to payload during BIRDIES-10 launch prevented recovery of sensor data. (4) All 7Be concentrations are decay-corrected to collection data (t = 0).
Range of measured 10Be/7Be from the precipitation campaign and likely moisture sources.
| Event | Measured 10Be/7Be | Hysplit Air Mass Origin |
|---|---|---|
| 2 | 2.09 (E32 anomaly) from system 1 | System 1: Pacific Ocean (>500 m a.g.l.); Gulf of Mexico (<500 m a.g.l.) |
| 3 | 1.74–1.95 | System 1: Pacific Ocean |
| 4 | 1.69–1.92 | Gulf of Mexico |
| 5 | 1.21–1.85 | System 1: Pacific Ocean |
| 6 | 1.22–2.05 | System 1 and 2: Pacific Ocean (>500 m a.g.l.); Gulf of Mexico (<500 m a.g.l.) |
References
1. Ehernberger, L.J. High Altitude Turbulence for Supersonic Cruise Vehicles; NASA Tech. Memo. 88285 NASA: Boulder, CO, USA, 1987; 15p.
2. Alexander, M.L.; Pfister, L. Gravity wave momentum flux in the lower stratosphere over convection. Geophys. Res. Lett.; 1995; 22, pp. 2029-2032. [DOI: https://dx.doi.org/10.1029/95GL01984]
3. Fritts, D.C.; Alexander, M.J. Gravity wave dynamics and effects in the middle atmosphere. Rev. Geophys.; 2003; 41, 1003. [DOI: https://dx.doi.org/10.1029/2001RG000106]
4. Hodges, R.R., Jr. Eddy diffusion coefficients due to instabilities in internal gravity waves. J. Geophys. Res.; 1969; 74, pp. 4087-4090. [DOI: https://dx.doi.org/10.1029/JA074i016p04087]
5. Lindzen, R.S. Turbulence and stress owing to gravity wave and tidal breakdown. J. Geophys. Res.; 1981; 86, pp. 9707-9714. [DOI: https://dx.doi.org/10.1029/JC086iC10p09707]
6. Andreassen, Ø.; Wasberg, C.-E.; Fritts, D.C.; Isler, J.R. Gravity wave breaking in two and three dimensions, 1, Model description and comparison of two-dimensional evolutions. J. Geophys. Res.; 1994; 99, pp. 8095-8108.
7. Lal, D.; Peters, B. Cosmic Ray Produced Radioactivities on the Earth. Handbuch der Physik; Sitte, K. Springer: Berlin/Heidelberg, Germany, 1967; pp. 551-612.
8. Lal, D.; Baskaran, M. Applications of Cosmogenic Isotopes as Atmospheric Tracers. Handbook of Environmental Isotope Geochemistry; Baskaran, M. Springer: Berlin/Heidelberg, Germany, 2011; pp. 575-589.
9. Raisbeck, G.M.; Yiou, F.; Fruneau, M.; Loiseaux, J.M.; Lieuvin, M.; Ravel, J.C.; Lorius, C. Cosmogenic 10Be concentrations in Antarctic ice during the past 30,000 years. Nature; 1981; 292, pp. 825-826. [DOI: https://dx.doi.org/10.1038/292825a0]
10. Gregory, J.C. Atmospheric radioactive isotopes at orbital altitudes. Radiat. Meas.; 1996; 26, pp. 841-850. [DOI: https://dx.doi.org/10.1016/S1350-4487(96)00092-3]
11. Koch, D.; Rind, D. Beryllium 10/beryllium 7 as a tracer of stratospheric transport. J. Geophys. Res.; 1998; 103, pp. 3907-3917. [DOI: https://dx.doi.org/10.1029/97JD03117]
12. Hidy, A.J.; Brown, T.A.; Tumey, S.J.; Zimmerman, S.R.H.; Gharibyan, N.; Frank, M.; Bench, G. A new 7 Be AMS capability established at CAMS and the potential for large datasets. Nucl. Instrum. Methods Phys. Res. B; 2018; 414, pp. 126-132. [DOI: https://dx.doi.org/10.1016/j.nimb.2017.11.017]
13. Peters, B. Radioactive beryllium in the atmosphere and on earth. Proc. Indian Acad. Sci. Sect. A; 1955; 41, pp. 67-71. [DOI: https://dx.doi.org/10.1007/BF03047175]
14. Poluianov, S.V.; Kovaltsov, G.A.; Mishev, A.L.; Usoskin, I.G. Production of cosmogenic isotopes 7Be, 10Be,14C, 22Na, and 36Cl in the atmosphere: Altitudinal profiles of yield functions. J. Geophys. Res.; 2016; 121, pp. 8125-8136. [DOI: https://dx.doi.org/10.1002/2016JD025034]
15. Nagai, H.; Tada, W.; Kobayashi, T. Production rates of 7Be and 10Be in the atmosphere. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. At.; 2000; 172, pp. 796-801. [DOI: https://dx.doi.org/10.1016/S0168-583X(00)00124-5]
16. Masarik, J.; Beer, J. Simulation of particle fluxes and cosmogenic nuclide production in the Earth’s atmosphere. J. Geophys. Res.; 1999; 104, pp. 12099-12111. [DOI: https://dx.doi.org/10.1029/1998JD200091]
17. Taylor, A.; Blake, W.H.; Iurian, A.R.; Millward, G.E.; Mabit, L. The Use of Be-7 as a Soil and Sediment Tracer. Assessing Recent Soil Erision Rates Through the Use of Beryllium-7 (Be-7); Mabit, L.; Blake, W. Springer: Cham, Switzerland, 2019; [DOI: https://dx.doi.org/10.1007/978-3-030-10982-0_1]
18. Jordan, C.E.; Dibb, J.E.; Finkel, R.C. 10Be/7Be tracer of atmospheric transport and stratosphere-troposphere exchange. J. Geophys. Res.; 2003; 108, 4234. [DOI: https://dx.doi.org/10.1029/2002JD002395]
19. Liu, N.; Liu, C.; Hayden, L. Climatology and detection of overshooting convection from 4 years of GPM precipitation radar and passive microwave observations. J. Geophys. Res.; 2020; 125, e2019JD032003. [DOI: https://dx.doi.org/10.1029/2019JD032003]
20. Homeyer, C.R.; Smith, J.B.; Bedka, K.M.; Bowman, K.P.; Wilmouth, D.M.; Ueyama, R.; Dean-Day, J.M.; St. Clair, J.M.; Hannun, R.; Hare, J. et al. Extreme Altitudes of Stratospheric Hydration by Midlatitude Convection Observed During the DCOTSS Field Campaign. Geophys. Res. Lett.; 2023; 50, e2023GL104914. [DOI: https://dx.doi.org/10.1029/2023GL104914]
21. Monaghan, M.C.; Krishnaswami, S.; Turekian, K.K. The global-average production of 10Be. Earth Planet. Sci. Lett.; 1986; 76, pp. 279-287. [DOI: https://dx.doi.org/10.1016/0012-821X(86)90079-8]
22. Brown, L.; Stensland, G.J.; Klein, J.; Middleton, R. Atmospheric deposition of 7Be and 10Be. Geochim. Cosmochim. Acta; 1989; 53, 135. [DOI: https://dx.doi.org/10.1016/0016-7037(89)90280-9]
23. Baskaran, M. A search for the seasonal variability on the depositional fluxes of 7Be and 210Pb. J. Geophys. Res.; 1995; 100, pp. 2833-2840. [DOI: https://dx.doi.org/10.1029/94JD02824]
24. Knies, D.L.; Elmore, D.; Sharma, P.; Vogt, S.; Li, R.; Lipshutz, M.E.; Petty, G.; Ferrel, J.; Monagham, M.C.; Fritz, S. et al. 7Be, 10Be, and 36Cl in precipitation. Nucl. Instr. Meth. Phys. Res.; 1994; 92, 340. [DOI: https://dx.doi.org/10.1016/0168-583X(94)96031-3]
25. Graham, I.; Ditchburn, R.; Barry, B. Atmospheric deposition of 7Be and 10Be in New Zealand rain (1996–1998). Geochem. Cosmochim. Acta; 2003; 67, pp. 361-373. [DOI: https://dx.doi.org/10.1016/S0016-7037(02)01092-X]
26. Heikkilä, U.; Beer, J.; Alfimov, V. Beryllium-10 and beryllium-7 in precipitation in Dübendorf (440 m) and at Jungfraujoch (3580 m), Switzerland (1998–2005). J. Geophys. Res.; 2008; 113, D11104. [DOI: https://dx.doi.org/10.1029/2007JD009160]
27. Skinner, S.; Butler, C.; Laurence, S.J.; Mangalam, A.S. In-Situ Measurements of Turbulent Velocity, Temperature, and Acoustic Fluctuations in the Stratosphere, AIAA 2020-0778. Proceedings of the American Institute of Aeronautics and Astronautics (AIAA) SciTech 2020 Forum; Orlando, FL, USA, 6–10 January 2020; [DOI: https://dx.doi.org/10.2514/6.2020-0778]
28. Zhu, W.; Skinner, S.; Laurence, S.J.; Wharton, S.; Hidy, A.J.; Ehrmann, T.S. Stratospheric Turbulence Measurements with Payload FiSH and Meteorological Balloons. Proceedings of the American Institute of Aeronautics and Astronautics (AIAA) Aviation and Aeronautics Forum; San Diego, CA, USA, 12–16 June 2023; [DOI: https://dx.doi.org/10.2514/6.2023-3544]
29. Cadet, D. Energy dissipation within intermittent clear air turbulence patches. J. Atmos. Sci.; 1977; 34, pp. 137-142. [DOI: https://dx.doi.org/10.1175/1520-0469(1977)034<0137:EDWICA>2.0.CO;2]
30. Theuerkauf, A.; Gerding, M.; Lübken, F.-J. LITOS—A new balloon-borne instrument for fine-scale turbulence soundings in the stratosphere. Atmos. Meas. Technol.; 2011; 4, pp. 55-66. [DOI: https://dx.doi.org/10.5194/amt-4-55-2011]
31. Faber, J.; Gerding, M.; Schneider, A.; Dornbrack, A.; Wilms, H.; Wagner, J.; Lubken, F.J. Evaluation of wake influence on high-resolution balloon-sonde measurements. Atmos. Meas. Technol.; 2019; 12, pp. 4191-4210. [DOI: https://dx.doi.org/10.5194/amt-12-4191-2019]
32. Lawrence, D.; Argrow, B.; Pointer, J.; Roseman, C.; Kenny, N.; Habeck, J.B.; Flaten, J.A.; Candler, G.V.; Barjatya, A.; Guardado, J. In-situ Turbulence and Particulate Measurements in Support of the BOLT II Flight Experiment. Proceedings of the American Institute of Aeronautics and Astronautics (AIAA) AIAA SciTech Form; San Diego, CA, USA, 12–16 June 2023; [DOI: https://dx.doi.org/10.2514/6.2023-0480]
33. Wallbrink, P.J.; Murray, A.S. Fallout of Be-7 in southeastern Australia. J. Environ. Radioact.; 1994; 25, pp. 213-228. [DOI: https://dx.doi.org/10.1016/0265-931X(94)90074-4]
34. Ioannidou, A.; Papastefanou, C. Precipitation scavenging of Be-7 and (CS)-C-137 radionuclides in air. J. Environ. Radioact.; 2006; 85, pp. 121-136. [DOI: https://dx.doi.org/10.1016/j.jenvrad.2005.06.005]
35. Bartholomew, M.J. Rain Gauge Instrument Handbook; Stafford, R. ARM Climate Research Facility Office of Biological and Environmental Research, U.S. Department of Energy: Washington, DC, USA, 2016; DOE/SC-ARM-TR-110 [DOI: https://dx.doi.org/10.2172/1245982]
36. Wharton, S.; Hidy, A.J.; Lee, J.; Repasch, M. Beryllium Isotopes in Rainfall for Stratospheric-Tropospheric Exchange Dynamics Experiment (BIRDIE) Field Campaign Report; U.S. Department of Energy, Atmospheric Radiation Measurement User Facility: Richland, WA, USA, 2023; DOE/SC-ARM-23-017
37. Available online: https://www.arm.gov/research/campaigns/sgp2022birdie (accessed on 1 November 2024).
38. Surano, K.A.; Hudson, G.B.; Failor, R.A.; Sims, J.M.; Holland, R.C.; MacLean, S.C.; Garrison, J.C. Helium-3 mass spectrometry for low-level tritium analysis of environmental samples. J. Radioanal. Nucl. Chem.; 1992; 161, pp. 443-453. [DOI: https://dx.doi.org/10.1007/BF02040491]
39. Zheng, M.; Liu, H.; Adolphi, F.; Muscheler, R.; Lu, Z.; Wu, M.; Prisle, N.L. Simulations of 7Be and 10Be with the GEOS-Chem global model v14.0.2 using state-of-the-art production rates. Geosci. Model Dev.; 2023; 16, pp. 7037-7057. [DOI: https://dx.doi.org/10.5194/gmd-16-7037-2023]
40. Herbst, K.; Muscheler, R.; Heber, B. The new local interstellar spectra and their influence on the production rates of the cosmogenic radionuclides 10Be and 14C. J. Geophys. Res.-Space Phys.; 2017; 122, pp. 23-34. [DOI: https://dx.doi.org/10.1002/2016JA023207]
41. Gordon, A.E.; Homeyer, C.R.; Smith, J.B.; Ueyama, R.; Dean-Day, J.M.; Atlas, E.L.; Smith, K.; Pittman, J.V.; Sayres, D.S.; Wilmouth, D.M. et al. Detailed examination of upper troposphere lower stratosphere composition change from DCOTSS airborne observations of active convection on 31 May 2022. EGUsphere; 2024; Preprint. [DOI: https://dx.doi.org/10.5194/egusphere-2023-2603]
42. Homeyer, C.R.; Bowman, K.P. Algorithm Description Document for Version 4.2 of the Three-Dimensional Gridded Nextrad WSR-88D Radar (GridRad) Dataset. Tech, rep. 2022; Available online: https://gridrad.org (accessed on 10 October 2022).
43. Schneider, A.; Wagner, J.; Faber, J.; Gerding, M.; Lübken, F.-J. Case study of wave breaking with high-resolution turbulence measurements with LITOS and WFR simulations. Atmos. Chem. Phys.; 2017; 17, pp. 7941-7954. [DOI: https://dx.doi.org/10.5194/acp-17-7941-2017]
44. Xuke, L.; Yunchong, F.; Li, Z.; Haogen, L.; Burr, G.S.; Yanting, B.; Guoqing, Z. Deep stratospheric intrusion events in China revealed on the ground by cosmogenic 10Be/7Be. Commun. Earth Environ.; 2024; 5, 553. [DOI: https://dx.doi.org/10.1038/s43247-024-01727-7]
45. Tian, C.; Wang, L.; Kaseke, K.F.; Bird, B.W. Stable isotope compositions (δ2H, δ18O and δ17O) of rainfall and snowfall in the central United States. Sci. Rep.; 2018; 8, 6712. [DOI: https://dx.doi.org/10.1038/s41598-018-25102-7]
46. Feely, H.W.; Larsen, R.J.; Sanderson, C.G. Factors that cause seasonal variations in Beryllium-7 concentrations in surface air. J. Environ. Radioact.; 1989; 9, pp. 223-249. [DOI: https://dx.doi.org/10.1016/0265-931X(89)90046-5]
47. Ioannidou, A.; Manolopoilou, M.; Papastefanou, C. Temporal changes of Be-7 and Pb-210 concentrations in surface air at temperate latitudes (40 degrees N). Appl. Radiat. Isot.; 2005; 63, pp. 277-284. [DOI: https://dx.doi.org/10.1016/j.apradiso.2005.03.010]
48. Aggarwal, P.K.; Romatschke, U.; Araguas-Araguas, L.; Belachew, D.; Longstaffe, F.J.; Berg, P.; Schumacher, C.; Funk, A. Proportions of convective and stratiform precipitation revealed in water isotope ratios. Nat. Geosci.; 2016; 9, pp. 624-629. [DOI: https://dx.doi.org/10.1038/ngeo2739]
49. Terzer-Wassmuth, S.; Araguás-Araguás, L.J.; Copia, L.; Wassenaar, L.I. High spatial resolution prediction of tritium (3H) in contemporary global precipitation. Sci. Rep.; 2022; 12, 10271. [DOI: https://dx.doi.org/10.1038/s41598-022-14227-5]
50. Stein, A.F.; Draxler, R.R.; Rolph, G.D.; Stunder, B.J.B.; Cohen, M.D.; Ngan, F. NOAA’s HYSPLIT Atmospheric Transport and Dispersion Modeling System. Bull. Am. Meteorol. Soc. (BAMS); 2015; 96, pp. 2059-2077. [DOI: https://dx.doi.org/10.1175/BAMS-D-14-00110.1]
51. Bowman, D.C.; Norman, P.E.; Pauken, M.T.; Albert, S.A.; Dexheimer, D.; Yang, X.; Krishnamoorthy, S.; Komjathy, A.; Cutts, J.A. Multihour stratospheric flights with the heliotrope solar hot-air balloon. J. Atmos. Ocean. Technol.; 2020; 37, pp. 1051-1066. [DOI: https://dx.doi.org/10.1175/JTECH-D-19-0175.1]
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Ehrmann, Thomas S 3 ; Zhu, Wenbo 4 ; Skinner, Shaun N 4 ; Beydoun, Hassan 1 ; Cameron-Smith, Philip J 1 ; Repasch, Marisa 5 ; Gunawardena, Nipun 1
; Lee, Jungmin M 1 ; Visser, Ate 6
; Griffin, Matthew 1 ; Maddren, Samuel 1 ; Oerter, Erik 6
1 Atmospheric, Earth and Energy Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA;
2 Atmospheric, Earth and Energy Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA;
3 Atmospheric, Earth and Energy Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA;
4 High-Speed Aerodynamic and Propulsion Laboratory, University of Maryland, College Park, MD 20742, USA;
5 Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA;
6 Nuclear and Chemical Science Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA;