Search NASA⌕ Search

SEARCH · Search NASA

Results for “Variable selection”

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 1,369 records · Page 76

Automatic detection of small bowel tumors in wireless capsule endoscopy images using ensemble learning

Purpose Wireless Capsule Endoscopy (WCE) is a minimally invasive diagnosis tool for lesion detection in the gastrointestinal tract, reaching places where conventional endoscopy is unable to. However, the significant amount of acquired data leads to difficulties in the diagnosis by the physicians; which can be eased with computer assistance. This paper addresses a method for the automatic detection of tumors in WCE by using a two‐step based procedure: region of interest selection and classification. Methods The first step aims to separate abnormal from normal tissue by using automatic segmentation based on a Gaussian Mixture Model (GMM). A modified version of the Anderson method for convergence acceleration of the expectation–maximization (EM) algorithm is proposed. The proposed features for both segmentation and classification are based on the CIELab color space, as a way of bypassing lightness variations, where the L component is discarded. Tissue variability among subjects, light inhomogeneities and even intensity differences among different devices can be overcome by using simultaneously features from both regions. In the second step, an ensemble system with partition of the training data with a new training scheme is proposed. At this stage, the gating network is trained after the experts have been trained decoupling the joint maximization of both modules. The partition module is also used at the test step, leading the incoming data to the most likely expert allowing incremental adaptation by preserving data diversity. Results This algorithm outperforms others based on texture features selected from Wavelets and Curvelets transforms, classified by a regular support vector machine (SVM) in more than 5%. Conclusions This work shows that simpler features can outperform more elaborate ones if appropriately designed. In the current case, luminance was discarded to cope with saturated tissue, facilitating the color perception. Ensemble systems remain an open research field. In the current case, changes in both topology and training strategy have led to significant performance improvements. A system with this level of performance can be used in current clinical practice.

Vieira, Pedro M.↗

Mediating Photochemical Reaction Rates at Lewis Acidic Rare Earths by Selective Energy Loss to 4f-Electron States

Manifesting chemical differences in individual rare earth (RE) element complexes is challenging due to the similar sizes of the tripositive cations and the corelike 4f shell. In this work, we disclose a new strategy for differentiating between similarly sized Dy 3+ and Y 3+ ions through a tailored photochemical reaction of their isostructural complexes in which the f-electron states of Dy 3+ act as an energy sink. Complexes RE(hfac) 3 (NMMO) 2 (RE = Dy (2-Dy) and Y (2-Y), hfac = hexafluoroacetylacetonate, and NMMO = N-methylmorpholine-N-oxide) showed variable rates of oxygen atom transfer (OAT) to triphenylphosphine under ultraviolet (UV) irradiation, as monitored by 1 H and 19 F NMR spectroscopies. Ultrafast transient absorption spectroscopy (TAS) identified the excited state(s) responsible for the photochemical OAT reaction or lack thereof. Competing sensitization pathways leading to excited-state deactivation in 2-Dy through energy transfer to the 4f electron manifold ultimately slows the OAT reaction at this metal cation. The measured rate differences between the open-shell Dy 3+ and closed-shell Y 3+ complexes demonstrate that using established principles of 4f ion sensitization may deliver new, selective modalities for differentiating the RE elements that do not depend on cation size.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Spectroscopic Studies of 30 Short-period Cataclysmic Variable Stars and Remarks on the Evolution and Population of Similar Objects

We present spectroscopy and orbital periods P {sub orb} for 30 apparently nonmagnetic cataclysmic binaries with periods below ∼3 hr, nearly all of which are dwarf novae, mostly of the SU Ursae Majoris subclass. We then turn to the evidence supporting the prediction that short-period dwarf novae evolve toward longer periods after passing through a minimum period—the “period bounce” phenomenon. Plotting data from the literature reveals that for superhump period excess ϵ = (P {sub sh} − P {sub orb})/P {sub orb} below ∼0.015, the period appears to increase with decreasing ϵ, agreeing at least qualitatively with the predicted behavior. Next, motivated by the long (decadal) outburst intervals of the WZ Sagittae subclass of short-period dwarf novae, we ask whether there could be a sizable population of “lurkers”—systems that resemble dwarf novae at minimum light, but which do not outburst over accessible timescales (or at all) and therefore do not draw attention to themselves. By examining the outburst history of the Sloan Digital Sky Survey sample of CVs, which were selected by color and not by outburst, we find that a large majority of the color-selected dwarf-nova-like objects have been observed to outburst, and we conclude that “lurkers,” if they exist, are a relatively minor part of the CV population.

79 ASTRONOMY AND ASTROPHYSICS↗

Analysis of Fault Permeability Using Mapping and Flow Modeling, Hickory Sandstone Aquifer, Central Texas

Reservoir compartments, typical targets for infill well locations, are commonly created by faults that may reduce permeability. A narrow fault may consist of a complex assemblage of deformation elements that result in spatially variable and anisotropic permeabilities. We report on the permeability structure of a km-scale fault sampled through drilling a faulted siliciclastic aquifer in central Texas. Probe and whole-core permeabilities, serial CAT scans, and textural and structural data from the selected core samples are used to understand permeability structure of fault zones and develop predictive models of fault zone permeability. Using numerical flow simulation, it is possible to predict permeability anisotropy associated with faults and evaluate the effect of individual deformation elements in the overall permeability tensor. We found relationships between the permeability of the host rock and those of the highly deformed (HD) fault-elements according to the fault throw. The lateral continuity and predictable permeability of the HD fault elements enhance capability for estimating the effects of subseismic faulting on fluid flow in low-shale reservoirs.

58 GEOSCIENCES↗

Is It Time to Tailor the Prediction of Radio-Induced Toxicity in Prostate Cancer Patients? Building the First Set of Nomograms for Late Rectal Syndrome

Purpose: Development of user-friendly tools for the prediction of single-patient probability of late rectal toxicity after conformal radiotherapy for prostate cancer. Methods and Materials: This multicenter protocol was characterized by the prospective evaluation of rectal toxicity through self-assessed questionnaires (minimum follow-up, 36 months) by 718 adult men in the AIROPROS 0102 trial. Doses were between 70 and 80 Gy. Nomograms were created based on multivariable logistic regression analysis. Three endpoints were considered: G2 to G3 late rectal bleeding (52/718 events), G3 late rectal bleeding (24/718 events), and G2 to G3 late fecal incontinence (LINC, 19/718 events). Results: Inputs for the nomogram for G2 to G3 late rectal bleeding estimation were as follows: presence of abdominal surgery before RT, percentage volume of rectum receiving >75 Gy (V75Gy), and nomogram-based estimation of the probability of G2 to G3 acute gastrointestinal toxicity (continuous variable, which was estimated using a previously published nomogram). G3 late rectal bleeding estimation was based on abdominal surgery before RT, V75Gy, and NOMACU. Prediction of G2 to G3 late fecal incontinence was based on abdominal surgery before RT, presence of hemorrhoids, use of antihypertensive medications (protective factor), and percentage volume of rectum receiving >40 Gy. Conclusions: We developed and internally validated the first set of nomograms available in the literature for the prediction of radio-induced toxicity in prostate cancer patients. Calculations included dosimetric as well as clinical variables to help radiation oncologists predict late rectal morbidity, thus introducing the possibility of RT plan corrections to better tailor treatment to the patient's characteristics, to avoid unnecessary worsening of quality of life, and to provide support to the patient in selecting the best therapeutic approach.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Synthesis, characterization and evaluation of nano-zirconium vanadate ion exchanger by using three different preparation techniques

Sol-gel, homogeneous precipitation and hydrothermal synthesis are three different preparation techniques have been used as an attempt to synthesize nano-zirconium vanadate with properties suitable to be used as ion exchangers. The impact of the synthetic preparation variables such as the reactant concentrations, reaction temperature and reaction time on the ion exchange capacity of the produced ion exchanger has been considered for each preparation technique. One sample from each preparation technique having the largest ion exchange capacity has been selected to be physically and chemically characterized using various analytical techniques such as XRD, TGA, DSC, pH titration, FTIR and SEM in order to determine the properties of the ion exchanger produced from each technique. For all the studied ZrV samples it can be presumed that they have the ion exchange affinity sequences for alkali metal ions K > Na > Li, the order for the alkaline earth metals is Ba > Ca > Mg and their affinity for radioactive metals follow Cs > Sr. Moreover, the prepared materials are of high thermal and radiation stabilities. Also they have high chemical stabilities toward wide concentration ranges of acid, basic as well as polar solvents. It has been deduced from the X-ray analysis that ZrV produced from the sol-gel technique has an amorphous structural. While those produced from the homogeneous precipitation and hydrothermal synthesis techniques, in the nano-scale have semi-crystalline structural. Furthermore, SEM confirms that particle size of the all studied prepared ZrV samples have nano-diameters of range 50-60 nm. Specific surface area of the three different prepared ion exchangers are found to be equal to 187, 192 and 320 m{sup 2}/g for sol-gel, homogeneous precipitation and hydrothermal, respectively. A tentative structural formula of Zr(OH){sub 2}(HVO{sub 4}){sub 2}.2H{sub 2}O has been proposed for all studied samples on the basis of on FTIR, DSC and TGA results.

36 MATERIALS SCIENCE↗

Suite of Activity-Based Probes for Cellulose-Degrading Enzymes

Microbial glycoside hydrolases play a dominant role in the biochemical conversion of cellulosic biomass to high-value biofuels. Anaerobic cellulolytic bacteria are capable of producing multicomplex catalytic subunits containing cell-adherent cellulases, hemicellulases, xylanases, and other glycoside hydrolases to facilitate the degradation of highly recalcitrant cellulose and other related plant cell wall polysaccharides. Clostridium thermocellum is a cellulosome producing bacterium that couples rapid reproduction rates to highly efficient degradation of crystalline cellulose. Herein, we have developed and applied a suite of difluoromethylphenyl aglycone, N-halogenated glycosylamine, and 2-deoxy-2-fluoroglycoside activity-based protein profiling (ABPP) probes to the direct labeling of the C. thermocellum cellulosomal secretome. These activity-based probes (ABPs) were synthesized with alkynes to harness the utility and multimodal possibilities of click chemistry, and to increase enzyme active site inclusion for LC-MS analysis. We directly analyzed ABP-labeled and unlabeled global MS data, revealing ABP selectivity for glycoside hydrolase (GH) enzymes in addition to a large collection of integral cellulosome-containing proteins. By identifying reactivity and selectivity profiles for each ABP, we demonstrate our ability to widely profile the functional cellulose degrading machinery of the bacterium. Derivatization of the ABPs, including reactive groups, acetylation of the glycoside binding groups, and mono- and disaccharide binding groups, resulted in considerable variability in protein labeling. Our probe suite is applicable to aerobic and anaerobic cellulose degrading systems, and facilitates a greater understanding of the organismal role associated within biofuel development.

Chauvigne-Hines, Lacie M↗

Cp* Noninnocence Leads to a Remarkably Weak C–H Bond via Metallocene Protonation

Metallocenes, including their permethylated variants, are extremely important in organometallic chemistry. In particular, many are synthetically useful either as oxidants (e.g., Cp 2 Fe + ) or as reductants (e.g., Cp 2 Co, Cp* 2 Co, and Cp* 2 Cr). The latter have proven to be useful reagents in the reductive protonation of small-molecule substrates, including N 2 . As such, understanding the behavior of these metallocenes in the presence of acids is paramount. In the present study, we undertake the rigorous characterization of the protonation products of Cp* 2 Co using pulse electron paramagnetic resonance (EPR) techniques at low temperature. We provide unequivocal evidence for the formation of the ring-protonated isomers Cp*(exo/endo-η 4 -C 5 Me 5 H)Co + . Variable temperature Q-band (34 GHz) pulse EPR spectroscopy, in conjunction with density functional theory (DFT) predictions, are key to reliably assigning the Cp*(exo/endo-η 4 -C 5 Me 5 H)Co + species. We also demonstrate that exo-protonation selectivity can be favored by using a bulkier acid and suggest this species is thus likely a relevant intermediate during catalytic nitrogen fixation given the bulky anilinium acids employed. Of further interest, we provide physical data to experimentally assess the C–H bond dissociation free energy (BDFE C–H ) for Cp*(exo-η 4 -C 5 Me 5 H)Co + . These experimental data support our prior DFT predictions of an exceptionally weak C–H bond (<29 kcal mol –1 ), making this system among the most reactive (with respect to C–H bond strength) to be thoroughly characterized. Furthermore, these data also point to the propensity of Cp*(exo-η 4 -C 5 Me 5 H)Co to mediate hydride (H – ) transfer. Our findings are not limited to the present protonated metallocene system. Accordingly, we outline an approach to rationalizing the reactivity of arene-protonated metal species, using decamethylnickelocene as an additional example.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Deep Learning for Subsurface Flow: A Comparative Study of U‐Net, Fourier Neural Operators, and Transformers in Underground Hydrogen Storage

Subsurface flow research is essential for the sustainable management of natural resources and the environment. Deep learning (DL) has significantly advanced this field by developing efficient and accurate surrogate models to replace computationally expensive physics‐based simulations. These surrogate models are commonly used to predict the spatiotemporal evolution of state variables, such as gas saturation and reservoir pressure, in heterogeneous geological formations. Despite the various DL models applied to this task, there is a lack of studies systematically comparing their performance. This absence of comparative analysis leads to somewhat arbitrary DL model selection in subsurface flow research, resulting in suboptimal performance and potentially inaccurate predictions. To bridge this gap, we conduct a systematic comparison study of three popular DL architectures—U‐Net, Fourier Neural Operators (FNO), and Segmentation Transformer (SETR)—in surrogate modeling of underground hydrogen storage (UHS). We focus on UHS due to its promise of enhancing clean energy resilience and its cyclic operational conditions that represent common scenarios in various subsurface applications. We evaluate the models based on accuracy, training cost, and inference speed. The comparison shows that U‐Net achieves the highest accuracy, followed by SETR and FNO. Despite its lower accuracy, FNO has the highest inference speed. SETR offers competitive accuracy with the least training memory usage, demonstrating the potential of transformers in learning subsurface flow. Our results provide guidance for selecting DL models for surrogate modeling in a wide range of subsurface flow problems.

42 ENGINEERING↗

Climate change drives convergent evolution of root traits on Sky Island climate relicts

Roots are essential to the strategies plants use to survive in variable environments, yet we know little of how they vary within species. Experimental conditions demonstrate that intraspecific plant root traits respond strongly to variation in the environment; however, it is unclear when these responses can be characterized as evolution in response to selective pressures of climate change over many generations. Sky Islands are model, natural climate relict ecosystems to examine climate-change driven evolution. Utilizing a common garden with replicate genotypes of Populus angustifolia (Narrowleaf cottonwood) from six Sky Island (SI) populations and nine adjacent Mountain Chain (MC) populations across three genetic provenances, we hypothesized that SI root traits have diverged due to historical isolation in warmer, drier climates. When grown in common conditions, populations originating on SI’s showed convergent evolution across three distinct genetic provenances, which was characterized by 44.16% decreased total root length, 42.64% decreased average root volume, 43.31% decreased root surface area, and significantly less root trait variation, relative to adjacent mountain chains. Convergent evolution of root traits from trees originating on SI’s is correlated with changes in mean annual precipitation and potential evapotranspiration in the field over the past ~ 125 years. These results demonstrate a consistent pattern in root trait evolution at the landscape scale and the role of climate on the evolution of root traits in a genetic and geographic context relevant to climate change.

Convergent evolution↗

A CHANDRA SURVEY OF THE X-RAY PROPERTIES OF BROAD ABSORPTION LINE RADIO-LOUD QUASARS

This work presents the results of a Chandra study of 21 broad absorption line (BAL) radio-loud quasars (RLQs). We conducted a Chandra snapshot survey of 12 bright BAL RLQs selected from Sloan Digital Sky Survey Data/Faint Images of the Radio Sky data and possessing a wide range of radio and C IV absorption properties. Optical spectra were obtained nearly contemporaneously with the Hobby-Eberly Telescope; no strong flux or BAL variability was seen between epochs. In addition to the snapshot targets, we include in our sample nine additional BAL RLQs possessing archival Chandra coverage. We compare the properties of (predominantly high-ionization) BAL RLQs to those of non-BAL RLQs as well as to BAL radio-quiet quasars (RQQs) and non-BAL RQQs for context. All 12 snapshots and 8/9 archival BAL RLQs are detected, with observed X-ray luminosities less than those of non-BAL RLQs having comparable optical/UV luminosities by typical factors of 4.1-8.5. (BAL RLQs are also X-ray weak by typical factors of 2.0-4.5 relative to non-BAL RLQs having both comparable optical/UV and radio luminosities.) However, BAL RLQs are not as X-ray weak relative to non-BAL RLQs as are BAL RQQs relative to non-BAL RQQs. While some BAL RLQs have harder X-ray spectra than typical non-BAL RLQs, some have hardness ratios consistent with those of non-BAL RLQs, and there does not appear to be a correlation between X-ray weakness and spectral hardness, in contrast to the situation for BAL RQQs. RLQs are expected to have X-ray continuum contributions from both accretion-disk corona and small-scale jet emission. While the entire X-ray continuum in BAL RLQs cannot be obscured to the same degree as in BAL RQQs, we calculate that the jet is likely partially covered in many BAL RLQs. We comment briefly on implications for geometries and source ages in BAL RLQs.

79 ASTRONOMY AND ASTROPHYSICS↗

A CONSTRAINT ON BROWN DWARF FORMATION VIA EJECTION: RADIAL VARIATION OF THE STELLAR AND SUBSTELLAR MASS FUNCTION OF THE YOUNG OPEN CLUSTER IC 2391

We present the stellar and substellar mass function (MF) of the open cluster IC 2391, plus its radial dependence, and use this to put constraints on the formation mechanism of brown dwarfs (BDs). Our multi-band optical and infrared photometric survey with spectroscopic follow-up covers 11 deg{sup 2}, making it the largest survey of this cluster to date. We observe a radial variation in the MF over the range 0.072-0.3 M {sub sun}, but no significant variation in the MF below the substellar boundary at the three cluster radius intervals is analyzed. This lack of radial variation for low masses is what we would expect with the ejection scenario for BD formation, although considering that IC 2391 has an age about three times older than its crossing time, we expect that BDs with a velocity greater than the escape velocity have already escaped the cluster. Alternatively, the variation in the MF of the stellar objects could be an indication that they have undergone mass segregation via dynamical evolution. We also observe a significant variation across the cluster in the color of the (background) field star locus in color-magnitude diagrams and conclude that this is due to variable background extinction in the Galactic plane. From our preliminary spectroscopic follow-up, to confirm BD status and cluster membership, we find that all candidates are M dwarfs (in either the field or the cluster), demonstrating the efficiency of our photometric selection method in avoiding contaminants (e.g., red giants). About half of our photometric candidates for which we have spectra are spectroscopically confirmed as cluster members; two are new spectroscopically confirmed BD members of IC 2391.

79 ASTRONOMY AND ASTROPHYSICS↗

A LIMIT ON THE NUMBER OF ISOLATED NEUTRON STARS DETECTED IN THE ROSAT ALL-SKY-SURVEY BRIGHT SOURCE CATALOG

Using new and archival observations made with the Swift satellite and other facilities, we examine 147 X-ray sources selected from the ROSAT All-Sky-Survey Bright Source Catalog (RASS/BSC) to produce a new limit on the number of isolated neutron stars (INSs) in the RASS/BSC, the most constraining such limit to date. Independent of X-ray spectrum and variability, the number of INSs is {<=}48 (90% confidence). Restricting attention to soft (kT{sub eff} < 200 eV), non-variable X-ray sources-as in a previous study-yields an all-sky limit of {<=}31 INSs. In the course of our analysis, we identify five new high-quality INS candidates for targeted follow-up observations. A future all-sky X-ray survey with eROSITA, or another mission with similar capabilities, can be expected to increase the detected population of X-ray-discovered INSs from the 8-50 in the BSC, to (for a disk population) 240-1500, which will enable a more detailed study of neutron star population models.

79 ASTRONOMY AND ASTROPHYSICS↗

Ensemble-Based Parameter Estimation in a Coupled GCM Using the Adaptive Spatial Average Method

Ensemble-based parameter estimation for a climate model is emerging as an important topic in climate research. And for a complex system such as a coupled ocean–atmosphere general circulation model, the sensitivity and response of a model variable to a model parameter could vary spatially and temporally. An adaptive spatial average (ASA) algorithm is proposed to increase the efficiency of parameter estimation. Refined from a previous spatial average method, the ASA uses the ensemble spread as the criterion for selecting “good” values from the spatially varying posterior estimated parameter values; these good values are then averaged to give the final global uniform posterior parameter. In comparison with existing methods, the ASA parameter estimation has a superior performance: faster convergence and enhanced signal-to-noise ratio.

54 ENVIRONMENTAL SCIENCES↗

Role of Smart Grids in Integrating Renewable Energy

This report was prepared for the International Smart Grid Action Network (ISGAN), which periodically publishes briefs and discussion papers on key topics of smart grid development globally. The topic of this report was selected by a multilateral group of national experts participating in ISGAN Annex 4, a working group that aims to produce synthesis insights for decision makers. This report is an update of a 2012 ISGAN Annex 4 report entitled “Smart Grid Contributions to Variable Renewable Resource Integration.” That report and other past publications of ISGAN Annexes can be found at www.iea-isgan.org and at www.cleanenergysolutions.org.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Climate variability and fill rate impacts on downstream flows from the Grand Ethiopian Renaissance Dam [Sheets]

Conclusions of this report include: Flow conditions (wet, normal, dry) are strong determinant of success in filling the dam; Some improvement for longer fill periods; Downstream Flow Compact only met under wetter flow conditions, rare otherwise; Results generated using other inflow datasets (not shown) indicate selection of data important for filling dam and highlight difficulty in determining which one to trust.

54 ENVIRONMENTAL SCIENCES↗

Policy-Driven Sustainable Saline Drainage Disposal and Forage Production in the Western San Joaquin Valley of California

Environmental policies to address water quality impairments in the San Joaquin River of California have focused on the reduction of salinity and selenium-contaminated subsurface agricultural drainage loads from westside sources. On 31 December 2019, all of the agricultural drainage from a 44,000 ha subarea on the western side of the San Joaquin River basin was curtailed. This policy requires the on-site disposal of all of the agricultural drainage water in perpetuity, except during flooding events, when emergency drainage to the River is sanctioned. The reuse of this saline agricultural drainage water to irrigate forage crops, such as ‘Jose’ tall wheatgrass and alfalfa, in a 2428 ha reuse facility provides an economic return on this pollutant disposal option. Irrigation with brackish water requires careful management to prevent salt accumulation in the crop root zone, which can impact forage yields. The objective of this study was to optimize the sustainability of this reuse facility by maximizing the evaporation potential while achieving cost recovery. This was achieved by assessing the spatial and temporal distribution of the root zone salinity in selected fields of ‘Jose’ tall wheatgrass and alfalfa in the drainage reuse facility, some of which have been irrigated with brackish subsurface drainage water for over fifteen years. Electromagnetic soil surveys using an EM-38 instrument were used to measure the spatial variability of the salinity in the soil profile. The tall wheatgrass fields were irrigated with higher salinity water (1.2–9.3 dS m -1 ) compared to the fields of alfalfa (0.5–6.5 dS m -1 ). Correspondingly, the soil salinity in the tall wheatgrass fields was higher (12.5 dS m -1 –19.3 dS m -1 ) compared to the alfalfa fields (8.97 dS mm -1 –14.4 dS mm -1 ) for the years 2016 and 2017. Better leaching of salts was observed in the fields with a subsurface drainage system installed (13–1 and 13–2). The depth-averaged root zone salinity data sets are being used for the calibration of the transient hydro-salinity computer model CSUID-ID (a one-dimensional version of the Colorado State University Irrigation Drainage Model). This user-friendly decision support tool currently provides a useful framework for the data collection needed to make credible, field-scale salinity budgets. In time, it will provide guidance for appropriate leaching requirements and potential blending decisions for sustainable forage production. This paper shows the tie between environmental drainage policy and the role of local governance in the development of sustainable irrigation practices, and how well-directed collaborative field research can guide future resource management.

54 ENVIRONMENTAL SCIENCES↗

Local Variability in PV Soiling Rate

We explore local and hyper-local variability in PV soiling rate by examining the correlation between PV system soiling rates over a multi-year period within two groups of First Solar soiling stations in separate regions of California. The results demonstrate significant seasonal variations in monthly-averaged soiling rates, with soiling varying from near 0 to ~0.1%/day in different months and with seasonal patterns that differ between sites. Monthly patterns at nearby sites were more highly correlated, suggesting that soiling rates can be strongly affected by local geography and weather. For the two regions studied, soiling rate pattern correlation diminished beyond a distance on the order of 50 km. In addition, for selected sites we demonstrate within-site variation in soiling rate of up to ~1.5-2x, which may be correlated with proximity to nearby agricultural activity.

14 SOLAR ENERGY↗