Search NASA⌕ Search

SEARCH · Search NASA

Results for “distribution shift”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Realized ecological forecast through an interactive Ecological Platform for Assimilating Data (EcoPAD, v1.0) into models

Abstract. Predicting future changes in ecosystem services is not only highly desirable but is also becoming feasible as several forces (e.g., available big data, developed data assimilation (DA) techniques, and advanced cyber-infrastructure) are converging to transform ecological research into quantitative forecasting. To realize ecological forecasting, we have developed an Ecological Platform for Assimilating Data (EcoPAD, v1.0) into models. EcoPAD (v1.0) is a web-based software system that automates data transfer and processing from sensor networks to ecological forecasting through data management, model simulation, data assimilation, forecasting, and visualization. It facilitates interactive data–model integration from which the model is recursively improved through updated data while data are systematically refined under the guidance of model. EcoPAD (v1.0) relies on data from observations, process-oriented models, DA techniques, and the web-based workflow. We applied EcoPAD (v1.0) to the Spruce and Peatland Responses Under Climatic and Environmental change (SPRUCE) experiment in northern Minnesota. The EcoPAD-SPRUCE realizes fully automated data transfer, feeds meteorological data to drive model simulations, assimilates both manually measured and automated sensor data into the Terrestrial ECOsystem (TECO) model, and recursively forecasts the responses of various biophysical and biogeochemical processes to five temperature and two CO 2 treatments in near-real time (weekly). Forecasting with EcoPAD-SPRUCE has revealed that mismatches in forecasting carbon pool dynamics are more related to model (e.g., model structure, parameter, and initial value) than forcing variables, opposite to forecasting flux variables. EcoPAD-SPRUCE quantified acclimations of methane production in response to warming treatments through shifted posterior distributions of the CH 4 :CO 2 ratio and the temperature sensitivity ( Q 10 ) of methane production towards lower values. Different case studies indicated that realistic forecasting of carbon dynamics relies on appropriate model structure, correct parameterization, and accurate external forcing. Moreover, EcoPAD-SPRUCE stimulated active feedbacks between experimenters and modelers to identify model components to be improved and additional measurements to be taken. It has become an interactive model–experiment (ModEx) system and opens a novel avenue for interactive dialogue between modelers and experimenters. Altogether, EcoPAD (v1.0) acts to integrate multiple sources of information and knowledge to best inform ecological forecasting.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

Deconstruction of the Ras switching cycle through saturation mutagenesis

Ras proteins are highly conserved signaling molecules that exhibit regulated, nucleotide-dependent switching between active and inactive states. The high conservation of Ras requires mechanistic explanation, especially given the general mutational tolerance of proteins. Here, we use deep mutational scanning, biochemical analysis and molecular simulations to understand constraints on Ras sequence. Ras exhibits global sensitivity to mutation when regulated by a GTPase activating protein and a nucleotide exchange factor. Removing the regulators shifts the distribution of mutational effects to be largely neutral, and reveals hotspots of activating mutations in residues that restrain Ras dynamics and promote the inactive state. Evolutionary analysis, combined with structural and mutational data, argue that Ras has co-evolved with its regulators in the vertebrate lineage. Overall, our results show that sequence conservation in Ras depends strongly on the biochemical network in which it operates, providing a framework for understanding the origin of global selection pressures on proteins.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1977

Utilization of the vacuum developed in the wake zone of space vehicles in the LDEF class

Contaminating fluxes from upper atmospheric molecules found in the wake of a typical spacecraft of the Long Duration Exposure Facility class are studied with the aim of determining vacuum levels which may be employed by experiments mounted on the vehicle. A simplified representation for the spacecraft, the velocity-shifted Maxwellian distribution (to evaluate ambient flux which has a velocity sufficient to overtake the spacecraft), and a calculation for the backscattered contribution to the flux density of the wake figure in the study are presented. Results indicate that direct and backscattered fluxes of about 10 to the seventh power molecules per sq cm can be expected in the wake of a spacecraft at 550 km altitude.

Oran, W. A.↗
NASA NTRS1994

Phase segregation due to simultaneous migration and coalescence

Ground-based modeling and experiments have been performed on the interaction and coalescence of drops leading to macroscopic phase separation. The focus has been on gravity-induced motion, with research also initiated on thermocapillary motion of drops. The drop size distribution initially shifts toward larger drops with time due to coalescence, and then a back towards smaller drops due to the larger preferentially settling out. As a consequence, the phase separation rate initially increases with time and then decreases.

Davis, Robert H.↗
NASA NTRS2001

Secondary metabolism in simulated microgravity

We have studied microbial secondary metabolism in a simulated microgravity (SMG) environment provided by NASA rotating-wall bioreactors (RWBs). These reactors were designed to simulate some aspects of actual microgravity that occur in space. Growth and product formation were observed in SMG in all cases studied, i.e., Bacillus brevis produced gramicidin S (GS), Streptomyces clavuligerus made beta-lactam antibiotics, Streptomyces hygroscopicus produced rapamycin, and Escherichia coli produced microcin B17 (MccB17). Of these processes, only GS production was unaffected by SMG; production of the other three products was inhibited. This was determined by comparison with performance in an RWB positioned in a different mode to provide a normal gravity (NG) environment. Carbon source repression by glycerol of the GS process, as observed in shaken flasks, was not observed in the RWBs, whether operated in the SMG or NG mode. The same phenomenon occurred in the case of MccB17 production, with respect to glucose repression. Thus, the negative effects of carbon source on GS and beta-lactam formation are presumably dependent on shear, turbulence, and/or vessel geometry, but not on gravity. Stimulatory effects of phosphate and the precursor L-lysine on beta-lactam antibiotic production, as observed in flasks, also occurred in SMG. An almost complete shift in the localization of produced MccB17 from cells to extracellular medium was observed when E. coli was grown in the RWB under SMG or NG. If a plastic bead was placed in the RWB, accumulation became cellular, as it is in shaken flasks, indicating that sheer stress favors a cellular location. In the case of rapamycin, the same type of shift was observed, but it was less dramatic, i.e., growth in the RWB under SMG shifted the distribution of produced rapamycin from 2/3 cellular:1/3 extracellular to 1/3 cellular:2/3 extracellular. Stress has been shown to induce or promote secondary metabolism in a number of other microbial systems. RWBs provide a low stress SMG environment, which, however, supports only poor production of MccB17, as compared to production in shaken flasks. We wondered whether the poor production in RWBs under SMG is due to the low level of stress, and whether increasing stress in the RWBs would raise the amount of MccB17 formed. We found that increasing shear stress by adding a single Teflon bead to the RWB improved MccB17 production. Although shear stress seems to have a marked positive effect on MccB17 production in SMG, addition of various concentrations of ethanol to RWBs (or to shaken flasks) failed to increase MccB17 production. Ethanol stress merely decreased production and, at higher concentrations, inhibited growth. Interestingly, cells growing in the RWB were much more resistant to the growth- and production-inhibitory effects of ethanol than cells growing in shaken flasks. With respect to S. hygroscopicus, addition of Teflon beads to the RWB reversed the inhibition of growth, but rapamycin production was still markedly inhibited, and the distribution did not revert back to a preferential cellular site.

NASA Discipline Cell Biology↗
NASA NTRS2018

Climate Change Impacts on Agriculture: Challenges, Opportunities, and AgMIP Frameworks for Foresight

Agricultural systems are currently undergoing rapid shifts owing to socioeconomic development, technological change, population growth, economic opportunity, evolving demand for commodities, and the need for sustainability amid global environmental change. It is not sufficient to maintain current harvest levels; rather, there is a need to rapidly increase production in light of a population growing to nearly 10 billion by mid-century and to more than 11 billion by 2100 (FAO, 2016; UN, 2016; Popkin et al., 2012). Current and future agricultural systems are additionally burdened by human-caused climate change, the result of accumulating greenhouse gas and aerosol emissions, ecological destruction, and land use changes that have altered the chemical composition of Earth’s atmosphere and trapped energy in the Earth system (IPCC, 2013; Porter et al., 2014). This increased energy has already raised average surface temperatures by approximately 1 degree Centigrade (GISTEMP Team, 2017; Hansen et al., 2010), leading early on to the term “global warming,” but this phenomenon is now more accurately referred to as “climate change” because it also modifies atmospheric circulation, adjusts regional and seasonal precipitation patterns, and shifts the distribution and characteristics of extreme events (Bindoff et al., 2013; Collins et al., 2013). Food and health systems face increasing risk owing to progressive climate change now manifesting itself as more frequent, severe extreme weather events—heat waves, droughts, and floods (IPCC, 2013). Often without warning, weather-related shocks can have catastrophic and reverberating impacts on the increasingly exposed global food system—through production, processing, distribution, retail, disposal, and waste. Simultaneously, malnutrition and ill health are arising from lack of access to nutritious food, exacerbated in crises such as food price spikes or shortages. For some countries, particularly import-dependent low-income countries, weather shocks and price spikes can lead to social unrest, famine, and migration.

Ruane, Alex C.↗
NASA NTRS2023

Quantification of Discharge-Specific Effects on Dissolved Organic Matter Export From Major Arctic Rivers From 1982 Through 2019

Long-term increases in Arctic river discharge have been well documented, and observations in the six largest Arctic rivers show strong positive correlations between dissolved organic carbon (DOC) concentration, river discharge, and chromophoric dissolved organic matter (CDOM) content. Here, observations of DOC and CDOM collected from 2009-2019 by the Arctic Great Rivers Observatory were used to estimate chromophoric DOC (CDOC) concentrations in the Kolyma, Lena, Mackenzie, Ob’, Yenisey, and Yukon Rivers. All rivers but the Mackenzie showed significant positive correlations between annual watershed runoff and the proportion of the DOC that is chromophoric. Historical estimates of DOC and CDOC export were calculated for 1982-2019 by extrapolating the DOC and CDOC concentration – discharge relationships from 2009-2019 as a hindcast modeled estimate. For the six rivers combined, modeled DOC and CDOC export increased, but CDOC increased faster than total DOC. The Lena and Ob’ Rivers showed significant increases in DOC export individually, with annual trends of 39.1 and 20.4 Gg C yr -1 respectively. November-April (winter) DOC and CDOC export increased in all rivers but the Yenisey, with the hindcast winter Kolyma export increasing by more than 20% per decade. There were no significant trends in discharge or associated DOC and CDOC fluxes during the observational period from 2009-2019; only when hindcasted values driven by changes in river discharge were analyzed did trends in DOC and CDOC emerge. This demonstrates how shifting seasonal distributions and increases in discharge can drive changes in DOC and CDOC concentrations and export, independent of other environmental factors.

carbon cycle↗
DOE OSTI2015

Metal-silicate catalysts: Single site, mesoporous systems without templates

The textural properties of a family of silicate and mixed metal-silicate materials prepared by a nonaqueous sol-gel reaction involving the cubic silicate Si 8 O 20 (SnMe 3 ) 8 and metal chlorides MCl 4 (M = Si, Ti, Zr) cross-linking reagents are described. Nitrogen adsorption isotherm data is presented and surface area and pore size distribution analyses for several examples of these materials are developed and correlated with the ratio of cross-linking reagent and the cubic silicate building block at the time of synthesis. Significant surface area and pore size distributions that shift to higher pore diameters are observed as the ratio of cross-linking reagent-to-cubic building block increases. A simple strategy for simultaneously controlling the porosity of these matrices while homogeneously dispersing identical metal centers on their surfaces for next generation catalysts is described.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Beyond Fine Tuning: Adding capacity to leverage few labels

In this paper we present a technique to train neural network models on small amounts of data. Current methods for training neural networks on small amounts of rich data typically rely on strategies such as fine-tuning a pre-trained neural networks or the use of domain-specific hand-engineered features. Here we take the approach of treating network layers, or entire networks, as modules and combine pre-trained modules with untrained modules, to learn the shift in distributions between data sets. The central impact of using a modular approach comes from adding new representations to a network, as opposed to replacing representations via fine-tuning. Using this technique, we are able surpass results using standard fine-tuning transfer learning approaches, and we are also able to significantly increase performance over such approaches when using smaller amounts of data.

data science↗
DOE OSTI2014

Monitoring Stock-Specific Abundance, Run Timing, and Straying of Chinook Salmon in the Columbia River Using Genetic Stock Identification (GSI)

Abstract Genetic stock identification (GSI) provides a method of characterizing stock-specific abundance, run timing, and straying as they pertain to the population demographics and behavior of both hatchery and wild fish and thus assists with fisheries management. In this study, tissues were collected and genotyped at 13 microsatellite loci from adult Chinook Salmon Oncorhynchus tshawytscha that passed Bonneville Dam during four consecutive migration years (2004–2007). GSI methods were used to estimate the proportions of stocks in a broad genetic baseline, with high-density coverage of Chinook Salmon across the eastern Pacific Rim. The estimates of abundance for interior Columbia River stocks showed widespread decline consistent with a shift to poor ocean conditions during the years when the fish in this study were at sea, but divergent stock- and age-specific patterns were also observed. For example, the upper Columbia River spring-run stock experienced a unique net gain in abundance during this period, and the interior Columbia River summer–fall-run stock experienced a delayed decline of 4-year-olds. Stocks with early run timing shifted to later migrations. However, run timing distributions did not shift unidirectionally for all stocks across years, and stock membership in three major run timing categories was maintained. Of 9,215 total adults sampled, 27 (0.3%) were out-of-basin strays and a quarter of the strays were putatively of wild origin. The general concordance of GSI results with those based on more traditional methods supports the effectiveness of GSI as a tool for fisheries management. Received May 5, 2013; accepted October 23, 2013 Published online February 10, 2014

Hess, Jon E.↗
NASA NTRS1992

Ventral horn cell responses to spaceflight and hindlimb suspension

Ventral horn cells of the lumbosacral enlargement of the spinal cord were studied focusing on the succinate dehydrogenase (SDH) activity and cross-sectional area of the soma of ventral horn cells which were measured using a computer-aided image-processing system. The relationships between the soma size and SDH activities of lumber ventral horn cells after 14 days of spaceflight (Cosmos 2044) or of hindlimb suspension are considered. Three groups of rats under consideration include control, 14-day spaceflight, and 14-day hindlimb suspension. Data obtained indicate that, compared to the control group, the population distribution of SDH activities in the flight rats shifted toward higher activities, whereas in the suspended rats the distribution shifted toward lower activities. The interactive effects within individual cells showed that there was a higher percentage of small cells with high SDH activities in the flight than in the control or suspended rats. It is suggested that changes in ventral horn cells were due to factors other than simply the absence of weight support.

Jiang, Bian↗
DOE OSTI2021

New Fracture Diagnostic Tool for Unconventionals: High-Resolution Distributed Strain Sensing via Rayleigh Frequency Shift during Production in Hydraulic Fracture Test 2

Fiber Optic monitoring in unconventional reservoirs has proven to be an invaluable diagnostic tool for assessing both near-wellbore stimulation effectiveness and to help describe the far-field frac geometries created by hydraulic fracture stimulation. Unfortunately, gaining any detailed qualitative and quantitative understanding of the near-wellbore frac geometry or cluster/stage productivity during production via Fiber Optic (FO) has proven to be more difficult, particularly in wells producing liquids. A new FO diagnostic method, Distributed Strain Sensing based on Rayleigh Frequency Shift (DSS-RFS), first demonstrated for oil and gas applications in the Hydraulic Test Site 2 (HFTS2) provides new insights about the characteristics of near-wellbore-region (NWR) during production. DSS-RFS is different from other FO strain measurements because it relies on accurate measurement of frequency shifts of Rayleigh backscattered spectrum obtained by scanning the fiber with a coherent optical time-domain reflectometer with a range of laser frequencies using a tunable-wavelength laser system. Changes in strain are measured with an extremely high spatial resolution of 20 cm and with high signal-to-noise ratios over long distances. In HFTS2, strain changes for the entire wellbore have been measured twice during scheduled shut-in and reopening operations (February 2020 and September 2020). After removing temperature effects, consistent strain changes have been observed at the location of most perforation clusters. These are caused by near wellbore fracture aperture changes due to pressure increases during shut-in within the near-wellbore fracture network. The strain-change patterns from the DSS-RFS during shut-in correlate very well with the location of clusters and allow for the definition of extending intervals with positive strain signals at each cluster and slightly compressing intervals with negative strain signals between the clusters and in the non stimulated intervals. The locations of the measured positive strain peaks also show good correspondence to DAS acoustic intensity measurements acquired during the stimulation. The geometry and magnitude of the strain changes differ significantly between the two tested completion designs in the same well. During shut in and reopening each cluster exhibit its own strain-change / pressure path. In addition, the September 2020 dataset also revealed the existence of small but measurable strain changes as consequence of pressure decline during production. These strain changes also correlate well with the presence of producing clusters, but the strain-rate signals are opposite to that obtained during shut-in and reopening operations. Although Downloaded from http://onepetro.org/URTECONF/proceedings-pdf/21URTC/2-21URTC/D021S031R002/2477551/urtec-2021-5408-ms.pdf/1 by Carol Worster on 28 February 2022 URTeC 5408 2 we are still in the early stages of exploring the potential of this novel FO technique, we believe that the highly detailed information contained in the measurement of strain changes using DSS-RFS during production can significantly improve our understanding of near-wellbore hydraulic fracture characteristics and the relationships between stimulation and production from unconventional oil and gas wells.

04 OIL SHALES AND TAR SANDS↗
DOE OSTI2025

A Data-Agnostic, Continuous Machine Learning Framework for Application in High Energy Physics and Beyond: Phase 1 Final Scientific/Technical Report

This Phase 1 effort has focused on the development of continual learning frameworks for use in machine learning, specifically in the applied context of High Energy Physics (HEP). Machine learning (ML) is a transformative technology by which computers, typically through the use of neural networks, are able to perform tasks with proficiency that rivals or surpasses that of human users. Model Degradation & Catastrophic Forgetting are two undesired phenomena which can occur in ML where the performance of a model degrades when either deployed on novel data streams, or trained on novel data which are sufficiently different than the data the models were initially trained on. A natural example where these sorts of effects can be observed is in the performance of detectors in harsh environments, where the detector signature may change over the lifetime of the detector as it ages and deteriorates — precisely what occurs in the experiments conducted in HEP. Real world HEP data is therefore an excellent test-ground and use-case for Continual Learning paradigms, which are techniques used in ML to counteract these problems. Ensemble learning is one such technique, where multiple smaller models are trained on subsets of the overall data and are ensembled together during inference. The intuition behind this technique is that, although there are shifts in the distributions which govern the incoming data streams, these shifts are not expected to be homogeneous or global. If a sufficient diversity in solutions within the various sub-models has been achieved, then at least one sub-model is expected to retain its performance within the overall ensemble. One further strength of this approach is that the architectures of the various models do not need to be identical, and in fact even different modalities of data can naturally be combined in this way. This work focused on applying ensemble learning techniques to derive results using two main datasets, anomaly detection in HEP data & time-series forecasting in semiconductor manufacturing data. Semiconductor manufacturing involves data with surprising similarity to that of HEP (e.g. wafer maps look very similar to digi-occupancy maps) and Cerium Lab’s prominence within the semiconductor industry makes semiconductor manufacturing a natural opportunity for commercialization of this work. Our efforts have led to two strong results. The first is that we evaluated the proposed ensembling techniques using previously proposed machine learning architectures for use in anomaly detection, namely AutoEncoder based models and their derivatives. We also developed new architectures which have not been evaluated in this context before. In fact, this work marks the first use of Vision Transformers for anomaly detection in HEP. Second, we demonstrated that ensemble learning significantly improves model performance in scenarios prone to degradation, validating its effectiveness across both HEP and semiconductor datasets. These results further support ensemble learning as a powerful strategy for mitigating catastrophic forgetting and maintaining robust performance in evolving data environments.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

The evolution of grain boundary energy in textured and untextured Ca–doped alumina during grain growth

The role of anisotropic grain boundary energy in grain growth is investigated using textured microstructures that contain a high proportion of special grain boundaries. Textured and untextured Ca–doped alumina was prepared by slip casting inside and outside a high magnetic field, respectively. At 1600°C, the textured microstructure exhibits faster growth than the untextured microstructure and its population of low–angle boundaries increases. Atomic force microscopy (AFM) is employed to measure the geometry of thermal grooves to assess the relative grain boundary energy of these systems before and after growth. In the textured microstructure, the grain boundary energy distribution narrows and shifts to a lower average energy. Conversely, the energy distribution broadens for the untextured microstructure as it grows and exhibits abnormal grain growth. Further analysis of the boundary networks neighboring abnormal grains reveals an energy incentive that facilitates their growth. These results suggest that coarsening is not the only dominant grain growth mechanism and that the system can lower its energy effectively by replacing high energy boundaries with those of low energy. The faster growth of lower energy boundaries suggests that isotropic simulations do not adequately account for anisotropic grain growth mechanisms or anisotropic mobility.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Measuring the Innermost Stable Circular Orbits of Supermassive Black Holes

We present a promising new technique, the g -distribution method, for measuring the inclination angle ( i ), the innermost stable circular orbit (ISCO), and the spin of a supermassive black hole. The g -distribution method uses measurements of the energy shifts in the relativistic iron line emitted by the accretion disk of a supermassive black hole due to microlensing by stars in a foreground galaxy relative to the g -distribution shifts predicted from microlensing caustic calculations. We apply the method to the gravitationally lensed quasars RX J1131–1231 ( z {sub s} = 0.658, z {sub l} = 0.295), QJ 0158–4325 ( z {sub s} = 1.294, z {sub l} = 0.317), and SDSS 1004+4112 ( z {sub s} = 1.734, z {sub l} = 0.68). For RX J1131−1231, our initial results indicate that r {sub ISCO} ≲ 8.5 gravitational radii ( r {sub g}) and i ≳ 55° (99% confidence level). We detect two shifted Fe lines in several observations, as predicted in our numerical simulations of caustic crossings. The current Δ E distribution of RX J1131–1231 is sparsely sampled, but further X-ray monitoring of RX J1131–1231 and other lensed quasars will provide improved constraints on the inclination angles, ISCO radii, and spins of the black holes of distant quasars.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Assessment of potential greenhouse gas mitigation from changes to crop root mass and architecture

Reducing (and eventually reversing) the increase in greenhouse gases (GHGs) in the atmosphere due to human activities, and thus reducing the extent and severity of anthropogenic climate change, is one of the great challenges facing humanity. While most of the man-caused increase in GHGs has been due to fossil fuel use, land use (including agriculture) currently accounts for about 25% of total GHG emissions and thus there is a need to include emission reductions from the land use sector as part of an effective climate change mitigation strategy. In addition, analyses included in the recent IPCC 5th Climate Change Assessment report suggests that it may not be possible to achieve large enough emissions reductions in the energy, transport and industrial sectors alone to stabilize GHG concentrations at a level commensurate with a less than 2°C global average temperature increase, without the help of a substantial CO 2 sink (i.e., atmospheric CO 2 removal) from the land use sector. One of the potential carbon sinks that could contribute to this goal is increasing C storage in soil organic matter on managed lands. This report details a preliminary scoping analysis, to assess the potential agricultural area in the US – where appropriate soil, climate and land use conditions exist – to determine the land area on which ‘improved root phenotype’ crops could be deployed and to evaluate the potential long-term soil C storage, given a set of ‘bounding scenarios’ of increased crop root input and/or rooting depth for major crop species (e.g., row crops (corn, sorghum, soybeans), small grains (wheat, barley, oats), and hay and pasture perennial forages). The enhanced root phenotype scenarios assumed 25, 50 and 100% increase in total root C inputs, in combination with five levels of modifying crop root distributions (i.e., no change and four scenarios with increasing downward shift in root distributions). We also analyzed impacts of greater root production on the soil-crop nitrogen balance, from the standpoint of increased need for additional N inputs and consequences for increased N 2 O flux, as well as potential impacts if more and deeper roots contributed to reduced N leaching. In the enhanced root phenotype scenarios, the implicit assumption was that increases in overall plant production could be achieved (e.g., through increased CO 2 assimilation, greater growth efficiency) without reducing the harvested yield – that is, we did not include potential leakage and land substitution effects from potential decreased crop yield in the analysis.

54 ENVIRONMENTAL SCIENCES↗