Current diagnostic velocity resolution limits our ability to search for exoplanets within the habitable zone. We propose a new capability (Crossfading-EDI or X-EDI) that will boost optical spectrograph stability and spectral resolution, enabling 100x-1000x more sensitivity for the exoplanet search. This technique for increasing the stability of any dispersive (grating or prism) spectrograph to unwanted wavelength drifts has been simulated on real data and shown to produce a 1000x or more stability improvement, by use of an interferometer using *pairs* of overlapping delays placed in series with the disperser, rather than singly-used delays. This combines fringe shifts having opposite phase reaction, to cancel drift. This technique, “Crossfading Externally Dispersed Interferometry” (X-EDI) builds upon an earlier singly- delayed Doppler technique (EDI) demonstrated on a variety of telescopes including the 5-meter Hale telescope at Mt. Palomar. The single-delay EDI technique already affords a significant stability enhancement to a spectrograph, and has been used by others to discover exoplanets in 2006 and 2016. We expect that improving EDI technique further by the use of crossfading pairs of delays will dramatically improve the instrumental noise floor due to spectrograph focal point drifts or detector pixel misplacement, which can limit the detection of small exoplanets over long (months or years) time scales. The X-EDI has been simulated on EDI data on a ThAr lamp line measured at the Hale telescope*. A simulated drift insult was applied. The observed reaction to the line position drift was reduced 1000x. All spectrographs suffer drift insults of various kinds, and the X-EDI technique reduces the reaction to these by moving the fine wavelength determination from the dispersive spectrograph to the interferometer, which uses the symmetry of delay pairs to eliminate drift. This technique can greatly improve spectral precision and stability for (1) Doppler radial velocimetry, and (2) direct planet imaging using adaptive optics (such as the Gemini Planet Imager) that feeds a low resolution integral field multi-object spectrograph. *David J. Erskine, J. Astr. Tele. Instrum. Sys., 7(2):025006, June 2021. Prepared by LLNL under Contract DE- AC52-07NA27344.

Arsenic (As) pollutes large regions of Asia. Despite phytoremediation initiatives using hyperaccumulators to remove As from contaminated soil, farmers remain reluctant to employ such strategies because of the low biomass and economic value of hyperaccumulating plants. In this study, we demonstrate that cassava can be used for As remediation using a high-throughput root phenotyping platform for cassava roots that we previously developed [1]. Using this phenotyping platform, we identified contrasting root traits associated with As uptake for the two genotypes Rayong 11(R11) and Rayong 90 (R90). Both cassava varieties were grown in pot systems under control (0 mg kg-1 As) and high As (50 mg kg-1 As) conditions and harvested 120 days after planting. We found As stress to reduce shoot and plant dry weight by 57% and 53%, respectively, whereas root dry weight and root traits showed only a slight change. Under As stress, R11 had a 75% higher nodal root number and a 59% lower basal root number than R90. Moreover, R11 root (100 mg kg-1 As) and branch (9 mg kg-1 As) tissues had considerably higher As concentrations than the same tissues in R90. The bioaccumulation coefficient for R11 (2.1) was significantly greater than for R90 (0.9). Additionally, bioethanol yields were unaffected by the presence of As in cassava starch. We suggest that cassava is a promising crop for phytoremediation and that root phenotyping is essential to breed cassava varieties with enhanced As uptake.

Deep learning (DL) methods have transformed the way we extract plant traits – both under laboratory as well as field conditions. Evidence suggests that “well-trained” DL models can significantly simplify and accelerate trait extraction as well as expand the suite of extractable traits. Training a DL model typically requires the availability of copious amounts of annotated data; however, creating large-scale annotated dataset requires non-trivial efforts, time, and resources. This has become a major bottleneck in deploying DL tools in practice. Self-supervised learning (SSL) methods give exciting solution to this problem, as these methods use unlabeled data to produce pretrained models for subsequent fine-tuning on labeled data, and have demonstrated superior transfer learning performance on down-stream classification tasks. We investigated the application of SSL methods for plant stress classification using few labels. Plant stress classification is a fundamentally challenging problem in that (1) disease classification may depend on abnormalities in a small number of pixels, (2) high data imbalance across different classes, and (3) there are fewer annotated and available plant stress images than in other domains. We compared four different types of SSL methods on two different plant stress datasets. We report that pre-training on unlabeled plant stress images significantly outperforms transfer learning methods using random initialization for plant stress classification. In summary, SSL based model initialization and data curation improves annotation efficiency for plant stress classification tasks.

Understanding root traits is essential to improve water uptake, increase nitrogen capture and raise carbon sequestration from the atmosphere. However, high-throughput phenotyping to quantify root traits for deeper field-grown roots remain a challenge. Recently developed open-source methods use image-based 3D reconstruction algorithms to build 3D models of plant roots from multiple 2D images and can extract root traits and phenotypes. Most of these methods rely on automated image orientation (Structure from Motion)[1] and dense image matching (Multiple View Stereo) algorithms to produce a 3D point cloud or mesh model from 2D images. Until now it is not known how the performance of these methods compares to each other when applied to field-grown roots. We investigate commonly used open-source pipelines on a test panel of twelve contrasting maize genotypes grown in real field conditions in this comparison study [2-6]. We compare 3D point clouds in terms of number of points, computation time, and model surface density. This comparison study will provide insight into the performance of different open-source pipelines for maize root phenotyping, and illuminates trade-offs between 3D model quality and performance cost for future high-throughput 3D root phenotyping.

Observations with high vertical resolution have revealed that power spectra of horizontal wind and temperature fluctuations versus vertical wavenumber m have a universal shape with a steep slope in a high m range, approximately proportional to m^(-3). Several theoretical models explaining this spectral slope were proposed under an assumption of gravity wave (GW) saturation. However, little evidence has been obtained to show that these universal spectra are fully composed of GWs. To confirm the validity of this assumption, two kinds of m spectra are calculated using outputs from a GW-permitting high-top general circulation model. One is the spectra for GWs designated by fluctuations having total horizontal wavenumbers of 21–639. The other is the spectra of fluctuations unfiltered except extracting a linear trend in the vertical that were often analyzed in the observational studies. Comparison between the two shows that GWs dominate the observed spectra only in a higher m part of the steep slope, whereas disturbances other than GWs significantly contribute to its lower m part. Moreover, geographical distributions of the characteristic wavenumbers, slopes, and spectral densities of GW spectra are examined for several divided height regions of the whole middle atmosphere. It is shown that strong vertical shear below the zonal wind jets as well as the wave saturation is responsible for the formation of the steep slopes of GW spectra.

On 12 January 2020, Taal volcano, Philippines, erupted after 43 years of repose, affecting more than 500,000 people. Using interferometric synthetic aperture radar (InSAR) data, we present the complete pre- to post-eruption analyses of the deformation of Taal. We find that: 1) prior to eruption, the volcano experienced long-term deflation followed by short-term inflation, reflecting the depressurization-pressurization of its ~5 km depth magma reservoir; 2) during the eruption, the magma reservoir lost a volume of 0.531 +/- 0.004 km^3 while a 0.643 +/- 0.001 km^3 lateral dike was emplaced; and 3) post-eruption analyses reveal that the magma reservoir started recovery approximately 3 weeks after the main eruptive phase. We propose a conceptual analysis explaining the eruption and address why, despite the large volume of magma emplaced, the dike remained at depth. We also report the unique and significant contribution of InSAR data during the peak of the crisis.

The extent of the fault damage zone remains an outstanding challenge confounding attempts to assess rock mass physical and mechanical properties, the effects on landscape evolution and slope stability, and to delineate safe places for human occupation and infrastructure development. Quantifying the relationship between faulting and the spatial geometrical and mechanical characteristics of a rock mass controlled by faulting is difficult, mainly because of varying lithology and rock mass characteristics, the effects of topography and vegetation and local erosion of weaker rock mass. Recent technological developments including Unmanned Aerial Vehicles, terrestrial laser scanning, photogrammetry and point cloud analysis software tools greatly enhance our ability to investigate the issues using the Yarlung Tsangpo (YLTP) Fault of southern Tibet as a case study where ideal geological conditions exist to investigate the relationship. In this study, the procedures, investigation approaches, evidence and criteria for defining the threshold distance for damage zones of YLTP Fault of southern Tibet were studied quantitatively by combining the spatial variations of fracture density, rock mass strength, rockfall inventory and previous thermal evidence. The results have been compared with published data from the evidence of thermal effects related to the exactly the same fault and show a good match between internal thermal action and rock mass physical and mechanical properties controlled by the same faulting. The extent of threshold distance of damage zone of the YLTP Fault is estimated as 5.9±0.6km. Within the damage zone, fracture density and cohesion of the rock mass show power curve relations with distance from the YLTP Fault. The internal dynamic action of fault controls rock mass physical and mechanical properties in the study area. The fault first affects the characteristics of rock mass structures, and then the orientation of the rock structures influences the stability of slope leading to rockfall.

Marine or coastal wetlands that host a diverse variety of flora and fauna are unique and fragile as they are subjected to changing coastlines and undergo dynamic spatial shifts with respect to tidal movements. In India, the Coastal Regulation Zone (CRZ) notification aims at the conservation of coastal regions, under the Environment (Protection) Act, 1986, and regulates developmental and construction activities within the CRZ regions of marine wetlands, in addition to the coastal belt. Remote sensing techniques can be of great use in understanding if the implementation of the CRZ has helped to regulate the proliferation of settlements in the wetland system. In this study, remote sensing techniques along with machine learning classifiers have been used for detecting and quantifying the recent settlements that have been built in the zones regulated by the CRZ of the Vembanad wetland of Kerala. Three standard change detection pre-processing techniques were used over Linear Imaging Self-Scanning Sensor (LISS) IV imagery which was followed by classification using machine learning algorithms: Support Vector Machine (SVM), random forest, and Artificial Neural Network (ANN) to identify the built-up erected in the CRZ region between 2012 and 2018. Comparing the performance of these classifiers, the random forest model was found to have the highest overall accuracy of 96%. It was found that the total area of new built-up that were constructed between 2012 and 2018 in the CRZ regions of 48 villages, that span across Ernakulam, Kottayam and Alappuzha districts of Kerala is 149 hectares. This usage of change detection techniques aided by machine learning algorithms over high-resolution LISS IV imagery would help to evaluate the effectiveness of the CRZ notification over other marine wetlands in India.

This article is composed of three independent commentaries about the state of ICON principles (Goldman et al., 2021) in the Geomagnetism, Paleomagnetism, and Electromagnetism (GPE) section and discussion on the opportunities and challenges of adopting them. Each commentary focuses on a different topic: Global collaboration, reproducibility, data sharing and infrastructure; Inclusive equitable, and accessible science: Involvement, challenges, and support of early career, BIPOC, women, LGBTQIA+, and/or disabled researchers; Community engagement, citizen science, education, and stakeholder involvement. Data sharing practices and open repository use still varies strongly between GPE communities. Some have a long tradition of data sharing, others are only starting it. Globally, GPE leadership is strongly dominated by white males and diversity may increase through the creation of Science Equality Commissions. Improved global stakeholder involvement can increase research impacts and help fight inequalities. In all investigated topics we see promising beginnings but also recognize obstacles that include a lack of funding, a lack of understanding of diversity, and prioritizing short-term gain over long-term benefit. Nonetheless, we are hopeful that our community will embrace ICON science.

Ancient exhumed accretionary complexes are sometimes associated with high-pressure–low-temperature (HP–LT) metamorphic rocks, such as psammitic schists, which are derived from sandy trench-fill sediment. At accretionary margins, sandy trench-fill sediments are rarely subducted to the depth of HP metamorphism because they are commonly scraped off at the frontal wedge. This study uses sandbox analogue experiments to investigate the role of seafloor topography in the transport of trench-fill sediment to depth during subduction. The experiments were conducted with a detached, rigid backstop to allow a topographic high (representing a seamount) to be subducted through a subduction channel. In experiments without topographic relief, progressive thickening of the accretionary wedge pushed the backstop down, leading to a stepping down of the décollement, narrowing the subduction channel, and underplating the wedge with subducting sediment. In contrast, in experiments with a topographic high, the subduction of the topographic high raised the backstop, leading to a stepping up of the décollement and widening of the subduction channel. These results suggest that the subduction of topographic relief is a possible mechanism for the transport of trench-fill sediment from the trench to HP environments through a subduction channel. A sufficient supply of sediment to the trench and topographic relief on the subducting oceanic plate might enable trench-fill sediment to be accreted at various depths and deeply subducted to become the protoliths of HP–LT metamorphic rocks.

Crop pest detection and mitigation remains an extremely challenging task for the farmers. Majority of the pest classification and detection techniques rely on supervised deep learning frameworks that require significant human intervention in labeling the input data, thereby making the down-stream tasks tedious. Therefore, this study presents a self-supervised learning (SSL) approach to classifying 12 types of agricultural insect pests from 9549 RGB images, by leveraging the Bootstrap your own latent (BYOL) algorithm. SSL uses minimal labeling and is indifferent to data augmentations or distortions. Hence, latent representations from pretrained SSL networks could be generalized well for downstream tasks like classification or object detection. For desirable classification of the insect images, the greatest challenges observed were: i) large intra-class variation (the same insect was found with different colors and patterns), and ii) complex background with inconspicuous foreground. Hence, to overcome these issues and aid generalizability of the representations learned through BYOL, entropy-guided segmentation (segments based on texture not color), is proposed as input to the SSL network in this study. Both raw and segmented images were separately fed to two independent BYOL SSL networks, i.e., with ResNet18 and ResNet50 architectures as the backbone. The efficacy of the latent representations for downstream applications was assessed using linear evaluation, and subsequently compared with traditional transfer learning outcomes from ResNet18 and ResNet50. The results indicated that, while ResNet50 backbone intuitively performed better in all cases, SSL aided with entropy-based segmentation led to ~94% classification accuracy compared to raw images (with ~90% maximum accuracy).

The development of new phenomics approaches to image and process data from the subcellular to ecosystem-scale has accelerated over the past decade. Many of these tools are produced “in-house” within a single lab or a group of collaborating labs, making it hard to keep up with the state-of-the-art for phenomics hardware and software development. The Plant Cell Atlas Phenomics Committee is creating a collaborative space that connects phenomics developers with each other and with the greater plant science community, with the goal of facilitating wide-reaching collaborations. To do this, we will be hosting a video series called “How We Built It” where developers provide a short tour of their inventions and relevant biological applications. We will follow the series with a more in-depth networking event where plant scientists can connect with the inventors and discuss collaborative opportunities. Our goal is to streamline the invention of new phenotyping tools and broaden the application of existing tools.

Understanding extreme space weather events is of paramount importance in efforts to protect technological systems in space and on the ground. Particularly in the thermosphere, the subsequent extreme magnetic storms can pose serious threats to low-Earth orbit (LEO) spacecraft by intensifying errors in orbit predictions. Extreme magnetic storms (minimum Dst $\leq$ –250 nT) are extremely rare: only 7 events occurred during the era of spacecraft with high-level accelerometers such as CHAMP (CHAllenge Mini-satellite Payload) and GRACE (Gravity Recovery And Climate experiment), and none with minimum Dst $\leq$ –500 nT, here termed magnetic superstorms. Therefore, current knowledge of thermospheric mass density response to superstorms is very limited. Thus, in order to advance this knowledge, four known magnetic superstorms in history, i.e., events occurring before CHAMP’s and GRACE’s commission times, with complete datasets, are used to empirically estimate density enhancements and subsequent orbital drag. The November 2003 magnetic storm (minimum Dst = –422 nT), the most extreme event observed by both satellites, is used as the benchmark event. Results show that, as expected, orbital degradation is more severe for the most intense storms. Additionally, results clearly point out that the time duration of the storm is strongly associated with storm-time orbital drag effects, being as important as or even more important than storm intensity itself. The most extreme storm-time decays during CHAMP/GRACE-like sample satellite orbits estimated for the March 1989 magnetic superstorm show that long-lasting superstorms can have highly detrimental consequences for the orbital dynamics of satellites in LEO.

Aerosols are postulated to alter moist convection by increasing cloud droplet number concentration. Cloud-resolving model simulations of radiative-convective equilibrium show that increased cloud droplet number concentration leads to higher convective mass flux, seemingly in line with a popular hypothesis which links the convective invigoration to delayed rain formation allowing more cloud liquid water to be frozen. Yet, the same phenomenon is also present in an alternative model configuration with only warm-rain microphysics, suggesting that one does not have to invoke ice microphysics. The key mechanism lies in the different vertical distributions of the increases in cloud liquid re-evaporation and in water vapor condensation, causing a dipole pattern that favors convection. This is further supported by a mechanism-denial experiment in which weakened cloud re-evaporation tends to mute invigoration. This work represents a major advancement of the process-level understanding of aerosol effects on convection.

Evidence for tectonic activity on Titan is provided by the presence of eroded mountain ranges. Xanadu is an equatorial region of Titan characterized by a complex topography, even though overall it has a lower average elevation compared to its surroundings. We investigated Xanadu’s southwestern margin, a part of the region which is comprised of heavily eroded and rugged terrains to the north and east and of smoother, more uniform terrains to the west and south. The central portions of southwestern Xanadu are characterized by an extensive fluvial network. The presence of such a distinctive feature was the main reason motivating the study of this area, given its potential to provide tectonic indications. Through detailed geomorphological mapping (map scale 1:700,000) on Synthetic Aperture Radar (SAR) data and analysis of both fluvial drainage patterns and Digital Terrain Models (DTMs), we identified several putative tectonic structures in this area: normal faulting to the west and east, thrust faulting to the north, small-scale strike-slip faulting in its central parts.Pull-apart basins are depressions bounded by both dip-slip faults and by (overlapping and/or bending) segments of a major transcurrent fault, i.e., they are basins generated in transtensional tectonic settings. We propose that central southwestern Xanadu is a pull-apart basin, bordered by both normal and thrust faults and formed by transtensional tectonics, which we consider to be the most recent tectonic phase active in this area. This basin is characterized by small-scale strike-slip faulting within it, on which a fluvial network has subsequently imposed.

The winter of 2019-20 was dominated by an extremely strong stratospheric polar vortex and positive tropospheric Arctic Oscillation (AO). Here, we analyze forecasts from 6 different prediction systems contributing to the C3S seasonal forecast database. Most performed very strongly, with consistently high skill for January-March 2020 from forecasts launched through October–December 2019. Although the magnitude of the anomalies was underestimated, the performance of most prediction systems was extremely high for a positive AO winter relative to the common hindcast climate. Ensemble members which better predicted the extremely strong stratospheric vortex better predicted the extreme tropospheric state. We find a significant relationship between forecasts of the anomalous mid-latitude tropospheric wave pattern in early winter, which destructively interfered with the climatological stationary waves, and the strength of the stratospheric vortex later in the winter. Our results demonstrate a strong interdependence between the accuracy of stratospheric vortex and AO forecasts.

Top-down estimates of CO2 fluxes are typically constrained by either surface-based or space-based CO2 observations. Both of these measurement types have spatial and temporal gaps in observational coverage that can lead to biases in inferred fluxes. Assimilating both surface-based and space-based measurements concurrently in a flux inversion framework improves observational coverage and reduces sampling biases. This study examines the consistency of flux constraints provided by these different observations and the potential to combine them by performing a series of six-year (2010–2015) CO2 flux inversions. Flux inversions are performed assimilating surface-based measurements from the in situ and flask network, measurements from the Total Carbon Column Observing Network (TCCON), and space-based measurements from the Greenhouse Gases Observing Satellite (GOSAT), or all three datasets combined. Combining the datasets results in more precise flux estimates for sub-continental regions relative to any of the datasets alone. Combining the datasets also improves the accuracy of the posterior fluxes, based on reduced root-mean-square differences between posterior-flux-simulated CO2 and aircraft-based CO2 over midlatitude regions (0.35–0.50~ppm) in comparison to GOSAT (0.39–0.57~ppm), TCCON (0.52–0.63~ppm), or in situ and flask measurements (0.45–0.53~ppm) alone. These results suggest that surface-based and GOSAT measurements give complementary constraints on CO2 fluxes in the northern extratropics and can be combined in flux inversions to improve observational coverage. This stands in contrast with many earlier attempts to combine these datasets and suggests that improvements in the NASA Atmospheric CO2 Observations from Space (ACOS) retrieval algorithm have significantly improved the consistency of space-based and surface-based flux constraints.

Reconstructions for global temperature development show an upward oscillation for the period of the 1880s through 1980s. This oscillation is being associated with natural variability and the temperature rise between the 1910s and 1940s with increased solar activity. The temperature impact of the 11-year solar cycle [Schwabe cycle] and the physical mechanism involved are insufficiently understood. Here, for the 22-year magnetic solar cycle [Hale cycle] a seawater surface temperature [SST] impact is described of 0,215 °C (0,238 ± 0,05 °C per W/m2); the derived impact for the 11-year cycle is 0,122 °C (0,135 ± 0,03 °C per W/m2). Also, a parallel development is described for seawater surface temperature [HadSST3 dataset] and the minima of total solar irradiance [LISIRD dataset] after a correction based on the 22-year solar cycle polarity change. With the correction, the combination of the positive and negative minima shows for the period 1890-1985 a high SST solar sensitivity: 1,143 ± 0,23 °C per W/m2 (with 90,5% declared variance). This implies that the Sun has caused a warming of 1,07 °C between Maunder minimum (late 17th century) and the most recent solar minimum year 2017 - which is well over half of the intermediate temperature rise of approximately 1,5 °C. The results demonstrate that the 22-year cycle forms a crucial factor required for better understanding the Sun-temperature relation. Ignoring the 22-year cycle leads to significant underestimation of the Sun’s influence in climate change combined with an overestimation of the impact of anthropogenic factors and greenhouse gases such as CO2.