Explore the vast range of titles available.

Efficiently navigating a superpressure balloon in the stratosphere 1 requires the integration of a multitude of cues, such as wind speed and solar elevation, and the process is complicated by forecast errors and sparse wind measurements. Coupled with the need to make decisions in real time, these factors rule out the use of conventional control techniques 2 , 3 . Here we describe the use of reinforcement learning 4 , 5 to create a high-performing flight controller. Our algorithm uses data augmentation 6 , 7 and a self-correcting design to overcome the key technical challenge of reinforcement learning from imperfect data, which has proved to be a major obstacle to its application to physical systems 8 . We deployed our controller to station Loon superpressure balloons at multiple locations across the globe, including a 39-day controlled experiment over the Pacific Ocean. Analyses show that the controller outperforms Loon’s previous algorithm and is robust to the natural diversity in stratospheric winds. These results demonstrate that reinforcement learning is an effective solution to real-world autonomous control problems in which neither conventional methods nor human intervention suffice, offering clues about what may be needed to create artificially intelligent agents that continuously interact with real, dynamic environments. Data augmentation and a self-correcting design are used to develop a reinforcement-learning algorithm for the autonomous navigation of Loon superpressure balloons in challenging stratospheric weather conditions.

DOGE cuts to the National Weather Service are threatening weather balloon launches. The Post's Scott Dance explains what that means for severe weather forecasts.

The suspected spy balloon was 60 metres tall, carrying a payload weighing around one tonne. By far the majority are weather balloons: these are launched twice a day simultaneously from almost 900 locations worldwide, according to the US National Weather Service. The US Federal Aviation Administration doesn't require tracking devices for payloads under 5.4 kilograms, or for launches or flight paths for such loads to be declared.

In just the past 5 years, the field of Earth observation has progressed beyond the offerings of conventional space-agency-based platforms to include a plethora of sensing opportunities afforded by CubeSats, unmanned aerial vehicles (UAVs), and smartphone technologies that are being embraced by both for-profit companies and individual researchers. Over the previous decades, space agency efforts have brought forth well-known and immensely useful satellites such as the Landsat series and the Gravity Research and Climate Experiment (GRACE) system, with costs typically of the order of 1 billion dollars per satellite and with concept-to-launch timelines of the order of 2 decades (for new missions). More recently, the proliferation of smartphones has helped to miniaturize sensors and energy requirements, facilitating advances in the use of CubeSats that can be launched by the dozens, while providing ultra-high (3–5 m) resolution sensing of the Earth on a daily basis. Start-up companies that did not exist a decade ago now operate more satellites in orbit than any space agency, and at costs that are a mere fraction of traditional satellite missions. With these advances come new space-borne measurements, such as real-time high-definition video for tracking air pollution, storm-cell development, flood propagation, precipitation monitoring, or even for constructing digital surfaces using structure-from-motion techniques. Closer to the surface, measurements from small unmanned drones and tethered balloons have mapped snow depths, floods, and estimated evaporation at sub-metre resolutions, pushing back on spatio-temporal constraints and delivering new process insights. At ground level, precipitation has been measured using signal attenuation between antennae mounted on cell phone towers, while the proliferation of mobile devices has enabled citizen scientists to catalogue photos of environmental conditions, estimate daily average temperatures from battery state, and sense other hydrologically important variables such as channel depths using commercially available wireless devices. Global internet access is being pursued via high-altitude balloons, solar planes, and hundreds of planned satellite launches, providing a means to exploit the internet of things as an entirely new measurement domain. Such global access will enable real-time collection of data from billions of smartphones or from remote research platforms. This future will produce petabytes of data that can only be accessed via cloud storage and will require new analytical approaches to interpret. The extent to which today's hydrologic models can usefully ingest such massive data volumes is unclear. Nor is it clear whether this deluge of data will be usefully exploited, either because the measurements are superfluous, inconsistent, not accurate enough, or simply because we lack the capacity to process and analyse them. What is apparent is that the tools and techniques afforded by this array of novel and game-changing sensing platforms present our community with a unique opportunity to develop new insights that advance fundamental aspects of the hydrological sciences. To accomplish this will require more than just an application of the technology: in some cases, it will demand a radical rethink on how we utilize and exploit these new observing systems.

The lack of measurements of three‐dimensional (3D) distribution of horizontal wind vectors is a major challenge in atmospheric science. Here, we develop an algorithm to retrieve winds for nine pressure levels at 1° grid spacing from 70°N to 70°S. The retrieval is done by tracking water vapor from the hyperspectral Cross‐track Infrared Sounder aboard two polar satellites (NOAA‐20 and Suomi‐NPP) that have overlapped tracks separated by 50 min. We impose a gross error check by flagging retrievals that are too different from ERA‐5 reanalysis. Testing the algorithm for the first week of January and July 2020 indicates that our algorithm yields 104 wind profiles per day and these 3D winds qualitatively agree with ERA‐5. Compared with radiosonde data, the errors are within the range of reported errors of cloud‐tracking winds. Plain Language Summary We developed an algorithm to derive winds in the tropics and midlatitudes across nine vertical levels from two polar‐orbiting satellites carrying high‐resolution sounding instruments. Our algorithm yields 104 wind profiles per day. Comparisons with measurements from weather balloons show that our wind retrievals are within the error range of existing satellite wind products (usually at one level at a given time and location) and outperform them in vertical resolution. Key Points Horizontal wind vectors are retrieved by tracking water vapor fields from an infrared sounder aboard two polar‐orbiting satellites Approximately 104 wind profiles are derived per day at nine levels from surface to 100 hPa at 1° resolution from 70°N to 70°S Compared to radiosonde data, retrieved winds have similar errors to cloud‐tracking winds

We describe and show results from a series of field campaigns that used balloonborne instruments launched from India and Saudi Arabia during the summers 2014–17 to study the nature, formation, and impacts of the Asian Tropopause Aerosol Layer (ATAL). The campaign goals were to i) characterize the optical, physical, and chemical properties of the ATAL; ii) assess its impacts on water vapor and ozone; and iii) understand the role of convection in its formation. To address these objectives, we launched 68 balloons from four locations, one in Saudi Arabia and three in India, with payload weights ranging from 1.5 to 50 kg. We measured meteorological parameters; ozone; water vapor; and aerosol backscatter, concentration, volatility, and composition in the upper troposphere and lower stratosphere (UTLS) region. We found peaks in aerosol concentrations of up to 25 cm–3 for radii > 94 nm, associated with a scattering ratio at 940 nm of ∼1.9 near the cold-point tropopause. During medium-duration balloon flights near the tropopause, we collected aerosols and found, after offline ion chromatography analysis, the dominant presence of nitrate ions with a concentration of about 100 ng m–3. Deep convection was found to influence aerosol loadings 1 km above the cold-point tropopause. The Balloon Measurements of the Asian Tropopause Aerosol Layer (BATAL) project will continue for the next 3–4 years, and the results gathered will be used to formulate a future National Aeronautics and Space Administration–Indian Space Research Organisation (NASA–ISRO) airborne campaign with NASA high-altitude aircraft.

Barometers floating on high‐altitude balloons in the relatively clement cloud layer on Venus could detect and characterize acoustic waves generated by seismic activity, avoiding the need for high‐temperature electronics required for surface seismology. Garcia et al. (2022, https://doi.org/10.1029/2022GL098844) recently demonstrated the detection of low‐frequency sound (infrasound) caused by earthquakes of magnitudes 7.3 and 7.5 from stratospheric balloons nearly 3,000 km away from the epicenter. They provided a preliminary demonstration of earthquake magnitude and location inversion, and the determination of S‐ and Rayleigh wave velocities using only their acoustic signature. Large earthquakes produce low‐frequency seismic waves that penetrate the interiors of planets; their detection at continental‐scale distances from a high‐vantage point demonstrates the feasibility of balloon‐based investigations of Venus' interior. We contextualize these results within the effort to perform seismology on Venus from balloons, discuss its limitations, and share perspectives on open research questions in this area. Plain Language Summary Scientists have peered inside Earth, the Moon, and Mars by recording how seismic waves pass through them. Venus' interior is a mystery, though, because the heat and pressure at its surface have thus far prevented direct measurement of seismic activity. Balloons with pressure recorders higher up in the Venusian atmosphere, where it is much cooler, can listen for the sounds the seismic waves make as a proxy for directly measuring ground motion. Before sending the balloons to Venus, it is important to make sure that this idea would actually work. The fact that (Garcia et al., 2022, https://doi.org/10.1029/2022GL098844) were able to record sound from large earthquakes thousands of kilometers away on Earth means that one should be able to do the same thing on Venus as well. While there are still details to work out, listening for quakes on Venus from balloons just got a major boost here on Earth. Key Points A network of stratospheric balloons captured infrasound from several large earthquakes This is the first detection of seismic activity from airborne sensors at continental distances A similar network on Venus could capture and geolocate seismic activity

Vertical profiles of methane (CH4) are essential for validating satellite observations and quantifying regional sources and sinks. This study presents the first two high‐resolution CH4 profiles (0–25 km) over southeastern China, an economically developed region, using AirCore measurements. The profiles exhibited distinct variations: CH4 increased from 25 to 15 km, remained stable (15–6 km), decreased sharply (6–3 km), then rose toward the surface (∼600 ppb range). While trends align with observations in northwest China, concentrations were higher. Wind patterns and balloon trajectories influenced the profiles, with long‐range air mass transport from coastal megacities elevating upper‐atmosphere CH4. Comparisons with TCCON, TROPOMI, and GOSAT‐2 revealed 26–39 ppb discrepancies in column‐averaged CH4, exposing resolution limitations and retrieval uncertainties. Pronounced day‐to‐day variability highlight influence from meteorological conditions and regional transport. These findings emphasize the need for higher spatiotemporal resolution monitoring to improve CH4 assessments and climate modeling.

Tropical tropopause layer (TTL) cirrus clouds play a key role in the Earth climate system. Yet the relative role of the various processes shaping them remains poorly known. Characterizing the temporal evolution of cloudy structures from observations is essential to address this issue but represents a challenge. Indeed, space‐ and airborne platforms move fast and mainly provide instantaneous snapshots. In boreal winter 2021–2022, two balloon‐borne lidars flew over the Equatorial Pacific Ocean, slowly drifting above the clouds. We use those unique nighttime observations to quantify the distribution of TTL cloud lifetime above this homogeneous region. This distribution is strongly asymmetric: half of the clouds live less than 1 hr, but their mean lifetime is about 6 hr. The few long‐lived clouds (>12 ${ >} 12$ hr) dominate the cloud cover. Those results compare reasonably well with TTL cirrus lifetimes in the ERA5 reanalysis, although the modeled TTL cloud cover is largely underestimated.