Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Sample size determination”

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 91 records · Page 5

Empirical data on the origin of 'new' comets

The mass distribution of a group of 62 comets from the sample of Marsden et al (1978) was determined by assuming that the size of a comet is in proportion to the surface area of its vaporizing nucleus. A bimodal distribution was found, with a gap between magnitudes 8.1 and 9. The 50 brighter comets (magnitudes 1 to 8) have been tentatively identified as true 'new' comets; their size distribution cannot be fit to a constant slope on a log-log graph. These comets were apparently formed by a mechanism which differs from the fragmentation process. The 12 fainter may be fragments of comets which split or dimmed during their first passages. Results indicate that new comets could be identified with pristine planetesimals; they decay rapidly, probably by splitting when they enter the inner solar system.

Delsemme, A. H.↗

On the siting of noble gases in E-chondrites

Fractions of six E-chondrites were separated by density, grain size, and chemical resistance to determine the siting of noble gases. The samples were taken from the Qingzhen (E3), Indarch (E4), Abee and Saint Saveur (E4-5), and Yilmia and North West Forrest (E6) meteorites. The Ar-rich component of the E6s was concentrated in the enstatite-rich fraction. This subsolar component was resistant to HCl and HNO3 treatment, but could be partially dissolved by HF, implying that the subsolar component is located in the enstatite. The noble gases were transported there by metamorphism. Xe-129 was found in the same regions in the E6s, but was in areas associated with chondrules in the E4s. Additionally, the carbon-rich fraction of the E4 sample displayed Xe and Ne/CCF-Xe isotopic ratios similar to that found in C-chondrites. E3 and E4 primordial gases were analogous, with no subsolar contribution.

Crabb, J.↗

Using radar ground-truth to validate and improve the location accuracy of a lightning direction-finding network

A technique is described in which isolated radar echoes associated with clusters of lightning strikes are used to validate and improve the location accuracy of a lightning-direction-finding network. Using this technique, site errors of a magnetic direction-finding network for locating lightning strikes to ground were accurately determined. The technique offers advantages over existing techniques in that large sample sizes are readily attainable over a broad area on a regular basis; the technique can also provide additional constraints to redundant data methods such as that described by Orville (1987). Since most lightning strike networks have either partial or full weather radar coverage, the technique is practical for all but a few users.

Goodman, Steven J.↗

Oxygen control of ethylene biosynthesis during seed development in Arabidopsis thaliana (L.) Heynh

An unforeseen side-effect on plant growth in reduced oxygen is the loss of seed production at concentrations around 25% atmospheric (50 mmol mol-1 O2). In this study, the model plant Arabidopsis thaliana (L.) Heynh. cv. 'Columbia' was used to investigate the effect of low oxygen on ethylene biosynthesis during seed development. Plants were grown in a range of oxygen concentrations (210 [equal to ambient], 160, 100, 50 and 25 mmol mol-1) with 0.35 mmol mol-1 CO2 in N2. Ethylene in full-sized siliques was sampled using gas chromatography, and viable seed production was determined at maturity. Molecular analysis of ethylene biosynthesis was accomplished using cDNAs encoding 1-aminocyclopropane-1-carboxylic acid (ACC) synthase and ACC oxidase in ribonuclease protection assays and in situ hybridizations. No ethylene was detected in siliques from plants grown at 50 and 25 mmol mol-1 O2. At the same time, silique ACC oxidase mRNA increased three-fold comparing plants grown under the lowest oxygen with ambient controls, whereas ACC synthase mRNA was unaffected. As O2 decreased, tissue-specific patterning of ACC oxidase and ACC synthase gene expression shifted from the embryo to the silique wall. These data demonstrate how low O2 modulates the activity and expression of the ethylene biosynthetic pathway during seed development in Arabidopsis.

NASA Discipline Plant Biology↗

Rock magnetic properties of the Arunta Block, Central Australia, and their implication for the interpretation of long-wavelength magnetic anomalies

Rock magnetic and petrologic studies of a suite of deep crustal rocks from the Arunta Block of Central Australia reveal that the granulite grade rocks are in general much more magnetic than the amphibolite grade samples irrespective of bulk rock composition. The dominant magnetic mineral in all samples is relatively pure magnetite as determined from thermomagnetic and electron microprobe analysis. The bulk magnetic properties are typical of pseudosingle-domain to multidomain size material. The samples from our study have very large remanences compared to previous crustal magnetic studies, with the granulites having a median natural remanent magnetization of 4.1 A/m and Koenigsberger ratio of 7.2. These remanences are relatively resistant to the thermal demagnetization, with nearly 50 percent of the magnetization remaining after 400 C demagnetization. Thus remanence may contribute significantly to the observed magnetic anomalies, including long-wavelength magnetic anomalies, the source of which resides at depth and therefore at elevated temperature, where a thermoviscous remanant magnetization along the present-day field is likely to dominate.

Kelso, Paul R.↗

Comet coma sample return instrument

The sample collection technology and instrument concept for the Sample of Comet Coma Earth Return Mission (SOCCER) are described. The scientific goals of this Flyby Sample Return are to return to coma dust and volatile samples from a known comet source, which will permit accurate elemental and isotopic measurements for thousands of individual solid particles and volatiles, detailed analysis of the dust structure, morphology, and mineralogy of the intact samples, and identification of the biogenic elements or compounds in the solid and volatile samples. Having these intact samples, morphologic, petrographic, and phase structural features can be determined. Information on dust particle size, shape, and density can be ascertained by analyzing penetration holes and tracks in the capture medium. Time and spatial data of dust capture will provide understanding of the flux dynamics of the coma and the jets. Additional information will include the identification of cosmic ray tracks in the cometary grains, which can provide a particle's process history and perhaps even the age of the comet. The measurements will be made with the same equipment used for studying micrometeorites for decades past; hence, the results can be directly compared without extrapolation or modification. The data will provide a powerful and direct technique for comparing the cometary samples with all known types of meteorites and interplanetary dust. This sample collection system will provide the first sample return from a specifically identified primitive body and will allow, for the first time, a direct method of matching meteoritic materials captured on Earth with known parent bodies.

Albee, A. L.↗

Particle size distribution of suspended solids in the Chesapeake Bay entrance and adjacent shelf waters

Characteristics of suspended solids, including total suspended matter, total suspended inorganics, total suspended organics, particle size distribution, and the presence of the ten most prominent particle types were determined. Four research vessels simultaneously collected samples along four transects. Samples were collected within a 2-hour period that coincided with the maximum ebb penetration of Chesapeake Bay outwelling. The distribution of primary and secondary particle size modes indicate the presence of a surface or near-surface plume, possibly associated with three sources: (1) runoff, (2) resuspension of material within the Bay, and/or (3) resuspension of material in the area of shoals at the Bay mouth. Additional supportive evidence for this conclusion is illustrated with ocean color scanner data.

Byrnes, M. R.↗

Heterogeneous reaction of ozone with aluminum oxide

Rates and collision efficiencies for ozone decomposition on aluminum oxide surfaces were determined. Samples were characterized by BET surface area, X-ray diffraction, particle size, and chemical analysis. Collision efficiencies were found to be between 2 times 10 to the -10 power and 2 times 10 to the -9 power. This is many orders of magnitude below the value of 0.000001 to 0.00001 needed for appreciable long-term ozone loss in the stratosphere. An activation energy of 7.2 kcal/mole was found for the heterogeneous reaction between -40 C and 40 C. Effects of pore diffusion, outgassing and treatment of the aluminum oxide with several chemical species were also investigated.

Keyser, L. F.↗

High-silica /greater than 60%/ lunar glasses in an Apollo 14 soil sample - Evidence for silicic lunar volcanism

The major-element compositions of 93 low-specific-gravity (less than 2.60) high-silica (greater than 60%) glass particles from a sample of lunar fines (14259,20) were determined by electron microprobe analyses. The size, shape, abundance, mineralogy, and major-element composition of more than 60% of the high-silica glasses is consistent with their being fragments of interstitial glass from mare basalts. However, one group of 30 glasses with between 72% and 78% SiO2 and an average of approximately 2.6% FeO can be distinguished from other high-silica glasses both chemically and petrographically. Glass particles with this composition do not contain crystalline inclusions and are fairly homogeneous not only within a single particle but also from particle to particle. The chemistry and petrology of these glasses suggest that they are not fragments of interstitial glass or shock-melted particles from a 'granitic' source rock. Rather, the homogeneity and lack of crystalline inclusions suggest that this group of high-silica glasses was the product of lunar acidic volcanism.

Glass, B. P.↗

Statistical methods for efficient design of community surveys of response to noise: Random coefficients regression models

Research studies of residents' responses to noise consist of interviews with samples of individuals who are drawn from a number of different compact study areas. The statistical techniques developed provide a basis for those sample design decisions. These techniques are suitable for a wide range of sample survey applications. A sample may consist of a random sample of residents selected from a sample of compact study areas, or in a more complex design, of a sample of residents selected from a sample of larger areas (e.g., cities). The techniques may be applied to estimates of the effects on annoyance of noise level, numbers of noise events, the time-of-day of the events, ambient noise levels, or other factors. Methods are provided for determining, in advance, how accurately these effects can be estimated for different sample sizes and study designs. Using a simple cost function, they also provide for optimum allocation of the sample across the stages of the design for estimating these effects. These techniques are developed via a regression model in which the regression coefficients are assumed to be random, with components of variance associated with the various stages of a multi-stage sample design.

Tomberlin, T. J.↗

Gas bubble detector

A gas bubble detector having a modulated IR source focused through a bandpass filter onto a venturi, formed in a sample tube, to illuminate the venturi with modulated filtered IR to detect the presence of gas bubbles as small as 0.01 cm or about 0.004 in diameter in liquid flowing through the venturi. Means are provided to determine the size of any detected bubble and to provide an alarm in the absence of liquid in the sample tube.

Mount, Bruce E.↗

The Remote Sensing of Mineral Aerosols and their Impact on Phytoplankton Productivity

The overall objective of this experiment was to test the iron hypothesis does the addition of iron to nutrient rich surface waters enhance productivity? Our specific objectives in this experiment included sampling and studying the marine aerosol size and type (which are related to chemical reactivity) during the PlumEx cruise to determine the importance of local (Galapagos Islands) versus long-range sources of atmospheric material. Detailed results of single particle analysis of our samples are being prepared for publication in two papers. We collect aerosol samples and they have been analyzed for trace metals and other elements. We are mapped aerosol distribution and the desert source areas around the Arabian Sea region. We did record a clear relationship between the aerosol radiance and synoptic weather patterns with distinct signals over the ocean northwest and southwest of Australia. While the interpretation was limited an aerosol climatology pattern was presented.

Tindale, Neil W.↗

The MSP 2001 Mars Environmental Compatibility Assessment (MECA)

The Mars Environmental Compatibility Assessment (MECA) will evaluate the Martian environment for soil and dust-related hazards to human exploration as part of the Mars Surveyor Program 2001 Lander. Sponsored by the Human Exploration and Development of Space (HEDS) enterprise, MECA's goal is to evaluate potential geochemical and environmental hazards that may confront future Martian explorers, and to guide HEDS scientists in the development of high fidelity Mars soil simulants. The integrated MECA payload contains a wet-chemistry laboratory, a microscopy station, an electrometer to characterize the electrostatics of the soil and its environment, and arrays of material patches to study the abrasive and adhesive properties of soil grains. The instrument will acquire soil samples with a robotic arm equipped with a camera. MECA will examine surface and subsurface soil and dust in order to characterize particle size, shape, hardness, and also physical characteristics that may provide clues to mineralogy. MECA will characterize soil/water mixtures with respect to pH, redox potential, total dissolved ions, and trace toxins. MECA will determine the nature of electrostatic charging associated with excavation of soil, and the influence of ionizing radiation on material properties. It will also observe natural dust accumulation on engineering materials. To accomplish these objectives, MECA is allocated a mass of 10 kg within an enclosure of 35 x 25 x 15 cm. The Wet Chemistry Laboratory (WCL) consists of four identical cells that will accept samples from surface and subsurface regions accessible to the Lander's robotic arm, mix them with water, and perform extensive analysis of the solution. Ion-selective electrodes and related sensors will evaluate total dissolved solids, redox potential, pH, and the concentration of many soluble ions and gases in wet Martian soil. These electrodes can detect potentially dangerous heavy-metal ions, emitted pathogenic gases, and the soil's corrosive potential. Experiments will include cyclic voltammetry and anodic stripping voltammetry. Complementary to the Viking experiments, the chemical laboratory will characterize the water-soil solution rather than emitted gases. Nonetheless, through analysis of dissolved gases it will be able to replicate many of the Viking observations related to oxidants. MECA's microscopy station combines optical and atomic-force microscopy (AFM) in an actively focused, controlled illumination environment to image particles from millimeters to nanometers in size. Careful selection of substrates allows controlled experiments in adhesion, abrasion, hardness, aggregation, magnetic and other properties. Special tools allow primitive manipulation (brushing and scraping) of samples. Soil particle properties including size, shape, color, hardness, adhesive potential (electrostatic and magnetic), will be determined using an array of sample receptacles and collection substrates. The simple, rugged atomic-force microscope will image in the submicron size range and has the capability of performing a particle-by-particle analysis of the dust and soil. On Earth, the earliest forms of life are preserved as microfossils. The atomic-force microscope will have the required resolution to image down to the scale of terrestrial microfossils and beyond. Mounted on the end of the robot arm, MECA's electrometer actually consists of four types of sensors: an electric field meter, several triboelectricity monitors, an ion gauge, and a thermometer. Tempered only by ultraviolet-light-induced ions and a low-voltage breakdown threshold, the dry, cold, dusty martian environment presents an imposing electrostatic hazard to both robots and humans. The field meter will measure the ambient field on nearby objects while the triboelectric sensors, using identical circuitry, will measure the charge accumulated on test substances as they are dragged through the soil by the arm. The ion chamber, open to the environment, will sense both charged dust and free ions in the air. Over and above the potential threat to electronics, the electrostatic environment holds one of the keys to transport of dust and, consequently, Martian meteorology. Viewed with the robot arm camera, the abrasion and adhesion plates are strategically placed to allow direct observation of the interaction between materials and soils on a macroscopic scale. Materials of graded hardness are placed directly under the robot arm scoop to sense wear and soil hardness. A second array, placed on the lander deck, is deployed after the dust plume of landing has settled. It can be manipulated in a primitive fashion by the arm, first having dirt deposited on it from the scoop and subsequently shaken clean. A third array will passively collect dust from the atmosphere. In addition to objectives related to human exploration, the MECA data set will be rich in information relevant to basic geology, paleoclimate, and exobiology issues. To understand both contemporaneous and ancient processes on Mars, the mineralogy, petrology, and reactivity of Martian surface materials should be constrained. The MECA experiment will shed light on these quantities through its combination of chemistry and microscopy. MECA will be capable of measuring the composition of ancient surface water environments, observing microscopic evidence of geological (and biological?) processes, inferring soil and dust transport, comminution and weathering mechanisms, and characterizing soil horizons that might be encountered during excavation.

Hecht, M. H.↗

The viability of photovoltaics on the Martian surface

The viability of photovoltaics (PV) on the Martian surface may be determined by their ability to withstand significant degradation in the Martian environment. Probably the greatest threat is posed by fine dust particles which are continually blown about the surface of the planet. In an effort to determine the extent of the threat, and to investigate some abatement strategies, a series of experiments were conducted in the Martian Surface Wind Tunnel (MARSWIT) at NASA Ames Research Center. The effects of dust composition, particle size, wind velocity, angle of attack, and protective coatings on the transmittance of light through PV coverglass were determined. Both initially clear and initially dusted samples were subjected both to clear winds and simulated dust storms in the MARSWIT. It was found that wind velocity, particle size, and angle of attack are important parameters affecting occlusion of PV surfaces, while dust composition and protective coatings were not. Neither induced turbulence nor direct current biasing up to 200 volts were effective abatement techniques. Abrasion diffused the light impinging on the PV cells, but did not reduce total coverglass transmittance by more than a few percent.

Gaier, James R.↗

Structure and kinematics of the broad-line regions in active galaxies from IUE variability data

IUE archival data are used here to investigate the structure nad kinematics of the broad-line regions (BLRs) in nine AGN. It is found that the centroid of the line-continuum cross-correlation functions (CCFs) can be determined with reasonable reliability. The errors in BLR size estimates from CCFs for irregularly sampled light curves are fairly well understood. BLRs are found to have small luminosity-weighted radii, and lines of high ionization tend to be emitted closer to the central source than lines of low ionization, especially for low-luminosity objects. The motion of the gas is gravity-dominated with both pure inflow and pure outflow of high-velocity gas being excluded at a high confidence level for certain geometries.

Koratkar, Anuradha P.↗

Identification of Martian Regolith Sulfur Components in Shergottites Using Sulfur K Xanes and Fe/S Ratios

Based on isotopic anomalies in Kr and Sm, Sr-isotopes, S-isotopes, XANES results on S-speciation, Fe/S ratios in sulfide immiscible melts [5], and major element correlations with S determined in impact glasses in EET79001 Lith A & Lith B and Tissint, we have provided very strong evidence for the occurrence of a Martian regolith component in some impact melt glasses in shergottites. Using REE measurements by LA-ICP-MS in shergottite impact glasses, Barrat and co-workers have recently reported conflicting conclusions about the occurrence of Martian regolith components: (a) Positive evidence was reported for a Tissint impact melt, but (b) Negative evidence for impact melt in EET79001 and another impact melt in Tissint. Here, we address some specific issues related to sulfur speciation and their relevance to identifying Martian regolith components in impact glasses in EET79001 and Tissint using sulfur K XANES and Fe/S ratios in sulfide immiscible melts. XANES and FE-SEM measurements in approx. 5 micron size individual sulfur blebs in EET79001 and Tissint glasses are carried out by us using sub-micron size beams, whereas Barrat and coworkers used approx. 90 micron size laser spots for LA- ICP-MS to determine REE abundances in bulk samples of the impact melt glasses. We contend that Martian regolith components in some shergottite impact glasses are present locally, and that studying impact melts in various shergottites can give evidence both for and against regolith components because of sample heterogeneity.

Sutton, S. R.↗

Applicability of Micro X-Ray Fluorescence Spectroscopy to Astromaterials Curation and Research

Introduction: The Astromaterials Acquisition and Curation Office at NASA’s Johnson Space Center (JSC) curates NASA’s astromaterial sample collections which includes: Apollo samples, Luna samples, Ant-arctic meteorites, cosmic dust particles, microparticle impacts into space-flown materials, Genesis solar wind atoms, Stardust comet Wild-2 particles, Stardust inter-stellar particles, Hayabusa asteroid Itokawa particles, Hayabusa 2 asteroid Ryugu particles, and future OSIRIS-Rex asteroid Bennu particles (landing in Sep-tember, 2023) [1–3]. To enhance JSC’s advanced cu-ration capabilities, we have recently installed a high-performance micro-X-ray fluorescence (µXRF) spec-trometer to assist in sample characterization through rapid, non-destructive, in-situ elemental analyses that do not require the sample preparation protocols (i.e., polishing and carbon-coating) commonly needed for electron beam analyses. With this new instrument, we are capable of detecting all elements down to carbon in a variable-pressure or He-purged chamber for anal-ysis of a wide range of sample types. Here we describe the instrumental set-up, capabilities, and applicability of µXRF analysis to astromaterials curation and re-search. Instrumentation and Methodology: The X-ray fluorescence and computed tomography lab (X-FaCT) lab at JSC is now equipped with a Bruker M4 Tornado Plus µXRF (Fig. 1). This system is an energy-dispersive x-ray spectrometer equipped with two 60 mm2 silicon drift detectors (SDD) that are able to be used simulta-neously for output count rates ~500,000 cps. New light element windows allow detecting and analyzing the entire elemental range from carbon to americium. Two x-ray tubes (micro-focus Rh with polycapillary lenses and W with collimators of 0.5, 1.0, 2.0, and 4.5 mm) with max excitation parameters of 50 kV, 30 W and 50 kV, 40 W, respectively, allow for more flexibility of the analysis of high energy lines. The motorized X-Y-Z stage has a mapping range of 190 x 160 mm and can support samples up to 7 kg (~15.5 lbs) and a height of 120 mm [4]. Analytical modes include elemental analysis (down to ~20 µm spot size) via point, line, or area of bulk materials (rock surfaces, thin sections, thick sections, etc.) as well as coating analysis (determination of thickness and composition) of samples. This system has a variable vacuum chamber (1 mbar to 1 atm) that is also equipped with a He-purge system which accommodates vacuum sensitive samples while still allowing detection of light elements at atmospheric pressure. Utility and Applicability of µXRF in Astro-materials research and exploration science (ARES): Elemental analysis using µXRF is commonly em-ployed for both terrestrial and planetary geological science disciplines [5]. It is especially useful for analy-sis of astromaterials given the limited sample prepara-tion required, which is not feasible for certain materi-als. Here we show select applications of µXRF anal-yses of astromaterials that can, have, and will be done at JSC’s X-FaCT lab. Point analysis: In-situ spot analyses (~20 µm spot size) on a cut slab of Martian meteorite NWA 10922 allowed for the discovery, qualitative elemental analy-sis and determination of different feldspar minerals [6]. These point analyses served as an effective prelim-inary step for subsequent quantitative analyses. Ana-lytical standards can be employed for more accurate quantification of µXRF spot analyses. Area analysis: This analytical mode measures all detectable elements (from C to Am) at each pixel (>5 µm pixel size) in a user-defined area. The results are shown as elemental maps which can be extracted as 16-bit TIFF’s for further data processing. In Fig. 2. we show elemental distribution maps of the high-Ti basalt 73001,531 that have been processed using ImageJ software. From these maps you can accurately and quickly (this map took ~50 mins.) identify mineral components, such as pyroxene, plagioclase, oxides, and phosphates, compositional zoning, and mineral textures. Detection of high-Z phases: µXRF techniques are es-pecially effective at analyzing trace minerals with high-atomic-number (high-Z) elements because the high-energy characteristic X-rays used (relative to SEM EDS) allow for mapping of K lines in elements up to La (typically SEM maps use L X-ray lines for elements >Zn, and these can often have interferences). Thus, µXRF is especially suited for identifying minerals like zircon, baddeleyite, REE-rich phosphates, Fe-rich met-als, oxides, sulfides, and phosphides [4]. In Fig. 3 we show elemental distribution maps for 73001,530 where we are able to correlate the original video image with, Zr, Si, Y and Hf elemental maps together identi-fying the location of a zircon. In this location you would expect lower Si compared to surrounding mate-rial, as well as higher Zr, Y, and Hf content compared to surrounding material, all of which is confirmed by our XRF ele-mental distribution maps (Figure 2.) Conclusions: The new M4 Tornado Plus µXRF within the Astromaterials Acquisition and Curation office at NASA JSC allows for rapid and non-destructive elemental analysis of astromaterials with limited or no sample preparation. µXRF analyses pro-vide crucial compositional knowledge for the prelimi-nary examination and curation of astromaterials. This instrument enhances the advanced curation capabili-ties in the X-FaCT laboratory at JSC by allowing pro-ductive, cohesive, and non-destructive multi-modal x-ray analyses on astromaterial samples, which is neces-sary for the comprehensive curation and study of our current and future astromaterial collections. Addition-ally, µXRF can provide complimentary information to researchers for studies on astromaterials. References: [1] Allen, C. et al., (2011). Chemie De Erde Geochemistry, 71, 1-20. [2] McCubbin, F. M. et al., (2016) 47th LPSC, abstract #2668 [3] Zeigler, R. A. et al., (2017) 48th LPSC, abstract #2772 [4] Bruker User Manual [5] Young et al., (2016) Appl. Geochemistry, 72, 77-87 [6] Mor-ris, R. V. et al., (2023) 54th LPSC.

E W O'Neal↗

Designing an Optimal Ensemble Strategy for GMAO S2S Forecast System

GMAO Sub/Seasonal prediction system (S2S) is being readied for a major upgrade to GEOS-S2S Version 3. An important factor in successful extended range forecast is the definition of an ensemble For initialization of the ensemble we propose a combination of lagged and burst initial conditions. We plan to run a relatively large ensemble of 40 members for sub-seasonal forecast (up to 3 months), at which point we sub-sample the ensemble, and continue the forecast with 10 members (up to 12 months). Here we present the results of the extensive testing of various ways to generate the perturbations to the initial conditions and the validation of the stratified sampling strategy we chose.To generate perturbations for the burst ensemble members we used scaled differences of pairs of analysis states separated by 1-10 days, randomly chosen from a corresponding season. We considered perturbing separately only the atmospheric fields or only the ocean or both of the forecast initial conditions. Considering varying separation times between the analysis states, we were able to produce perturbations sampling various modes of variability. Focusing on the ENSO SST indices, we found that all types of perturbations are important for the ensemble spread.Our ensemble size for sub-seasonal forecasts was determined as to maximize the skill of predicting some of the leading modes of boreal winter atmospheric modes, NAO, PNA and AO. It is not feasible to run equally large ensemble for seasonal forecasts. Using a stratified sampling procedure we can identify the emerging directions of error growth. By comparing the stratified ensemble with randomly sampled ensemble of the same size, we were able to show that the former better estimates the mean of the original large ensemble.

Borovikov, Anna↗