Laboratory Measurement Uncertainty Research Articles

To calibrate or not to calibrate, that is the question

Sensors used for control have become widespread in water resources recovery facilities during the strive for resource efficient operations. However, their accuracy is reliant on uncertain laboratory measurements, which are used for calibration and, in turn, to correct for sensor drift. At the same time, current sensor calibration practices are lacking clear theoretical understanding of how measurement uncertainties impact the final control action. The effects of a customarily, and ad hoc, applied calibration threshold are unknown, leading to the current situation where many wastewater treatment processes are controlled by measurements with unknown accuracy. To study how sensor accuracy is affected by calibration, including varying calibration thresholds, we developed a simple theoretical model with closed-form expressions based on the variance and bias in sensor and laboratory measurements. The model was then simulated to yield the results, which showed no practical gain of using a calibration threshold, apart from the situation when calibration is more time-consuming than validation. By contrast, the best accuracy was obtained when consistently executing calibration, which opposes common practice. Further, the sensor calibration error was shown to be transferred to the process, causing a similar deviation from the setpoint when the same sensor was used for control. This emphasizes the importance of minimizing laboratory measurement uncertainties during calibration, which otherwise directly impact operations. Due to these findings we strongly advice shifting mindset from considering calibration as a sequential detection and correction approach, towards an estimation approach, aiming to estimate bias magnitude and drift speed.

Uncertainties in the permittivity model for seawater in FASTEM and implications for the calibration/validation of microwave imagers

This paper evaluates uncertainties in the ocean permittivity model needed to compute surface emission and reflection, at frequencies of 6 - 90 GHz. The study transforms permittivities and laboratory measurement uncertainties into typical brightness temperatures for microwave imagers. It is established that the permittivity model used by FASTEM agrees with the laboratory measurements to within their published uncertainties at frequencies of 23 GHz, 37 GHz and 89 GHz, but this is not the case at 6.8, 10.65 and 18.7 GHz, particularly at low temperatures.The laboratory measurement uncertainties transformed into typical brightness temperatures are reported and compared to observation minus short-range forecast (O - B) differences for microwave imagers for a range of ocean surface skin temperatures. For GMI and AMSR-2 these differences show very little skin-temperature dependent bias when calculated using FASTEM. However, when a new version of the permittivity model is used in FASTEM, that fits the low-frequency laboratory data better, an increase in skin temperature dependent bias is observed for channels at 10 - 23 GHz that is larger than the stated measurement uncertainties.This suggests that either the laboratory measurements have additional sources of systematic error at low frequencies, or there are additional sources of bias in the mean O - B calculations for microwave imagers that are currently masked by biases in FASTEM. To resolve this, reference-quality measurements of seawater permittivity are needed at all relevant microwave frequencies and temperatures, particularly at low temperatures and low frequencies.

Results of four European round-robins on short-circuit current temperature coefficient measurements of photovoltaic devices of different size

Within the EURAMET ENG55 “PhotoClass” project, several characteristics of photovoltaic (PV) devices beyond their performance at Standard Test Conditions were investigated, including measurements at varying irradiance and temperature. Four groups of PV devices of different size and technology were prepared and corresponding round-robins were run between partner laboratories with substantially different facilities and methods – namely based on spectral or integral measurements. This paper presents the outcome of the four inter-laboratory comparisons dealing with temperature coefficient measurements of the short-circuit current of PV devices, from reference-cell size to full-size commercial modules of mainly several c-Si technologies, but also with some examples of CIGS and GaAs devices. The measurement results are compared via En number assessment, hence including measurement uncertainties. The main outcome of this measurement exercise is a very good agreement of all the laboratories although completely different approaches were applied. In some cases, laboratory measurement uncertainties are even considered rather conservative and could therefore be revised. Furthermore, a comparison between bare cells and commercial modules of the same technology is made, which may represent useful information for PV manufacturers.

Field-based soil-texture estimates could replace laboratory analysis

Texture is one of the most important and most frequently measured parameters in soil science. It is common knowledge among field experienced soil scientists that soil texture can be well estimated in the field manually with so called “texture-by-feel”. However, no systematic evaluation exists that assessed the precision and accuracy of field based texture estimates as compared to the common, but time consuming, standard laboratory methods. In the course of the German Agricultural Soil Inventory, the texture of 3896 soil samples from 728 soil pits was estimated manually in the field and measured in the laboratory using standard sedimentation techniques. The field based estimations of the sand, silt and clay content showed a relative deviation from the measurements of only 3.8, 11.5 and 15.5%, respectively. The absolute uncertainty of field texture was 23, 32 and 17gkg−1 for sand, silt and clay, respectively. A large fraction (57–72%) of deviations between field and laboratory derived texture estimates was due to the laboratory measurement uncertainty, and due to the fact that only texture classes were estimated in the field and not mass fractions. Our findings indicate that for most purposes it is sufficient to estimate the soil texture manually “by feel” instead of conducting expensive particle size analyses in the laboratory.

Final report on CCQM-P73: International comparison of nitrogen monoxide in nitrogen gas standards (30–70) µmol/mol

The pilot study evaluated the level of comparability of laboratories' preparative capabilities for gravimetric nitrogen monoxide/nitrogen primary reference mixtures in the range (30–70) µmol/mol. The comparison was designed so that measurements would be performed at a central laboratory (the BIPM) and measurement results compared to values assigned by each national metrology institute (NMI) based on gravimetry using regression analysis.The advantages and complications in organizing a comparison with measurements performed at a central laboratory have been clearly demonstrated, notably: analytical measurement uncertainties can be reduced; a degree of equivalence parameter and its uncertainty can be calculated; regression analysis and therefore reference values for the comparison are highly dependent on the data set chosen for regression analysis and the uncertainty ascribed to the analytical measurement system within the central laboratory.Participating laboratories reported standard uncertainties for gravimetric preparation which ranged from 0.01% to 0.22% relative to the nitrogen monoxide mole fraction value.Following FTIR analysis, the standards of three laboratories (six gas standards in total) were omitted from the regression analysis data set, as a significant difference between reported and measured values of impurity contents was observed. An additional standard was removed from the regression analysis set as its deviation from the regression line was of the same order of magnitude as the standards already omitted. A regression line consistent with the remaining calibration data (15 gas standards) could be obtained by increasing estimates of the coordinating laboratory's measurement uncertainty by approximately a factor of three, resulting in analytical standard uncertainties of 0.12% (at 70 µmol/mol) and 0.27% (at 30 µmol/mol), and predicted standard uncertainties of the nitrogen monoxide mole fractions of 0.09 µmol/mol. Reported standard uncertainties related to gravimetric preparation ranged from 0.004 µmol/mol to 0.11 µmol/mol with a median of 0.03 µmol/mol. Nevertheless, comparison of the results of CCQM-P73 to previous key comparisons for nitrogen monoxide in nitrogen at 100 µmol/mol (CCQM-K1.c and EUROMET.QM-K1.c) illustrates the reductions in uncertainties that can be achieved through a comparison with measurements performed at a central facility.Main text.To reach the main text of this paper, click on Final Report.The final report has been peer-reviewed and approved for publication by the CCQM-GAWG.

PM2.5 Speciation Trends Network: Evaluation of Whole-System Uncertainties Using Data from Sites with Collocated Samplers

The objectives of this paper are to contrast the relative variability of replicate laboratory measurements of selected chemical components of fine particulate matter (PM) with total variability from collocated measurements and to compare the magnitudes of the uncertainties determined from collocated sampler data with those currently being provided to U.S. Environmental Protection Agency (EPA)’s Air Quality System (AQS) database by RTI International (RTI). Pointwise uncertainty values are needed for modeling and data analysis and should include all the random errors affecting each data point. Total uncertainty can be decomposed into two primary components: analytical measurement uncertainty and sampling uncertainty. Analytical measurement uncertainties are relatively easy to calculate from routine quality control (QC) data. Sampling uncertainties, on the other hand, are comparatively difficult to measure. In this paper, the authors describe data from collocated samplers to provide a snapshot of whole-system uncertainty for several important chemical species. The components of uncertainty were evaluated for key species from each of the analytical methods employed by the PM2.5 Speciation Trends Network (STN) program: gravimetry, ion chromatography (IC), X-ray fluorescence (XRF), and thermal-optical analysis for organic carbon and elemental carbon. The results show that the laboratory measurement uncertainties are typically very small compared with uncertainties calculated from the differences between samples collected from collocated samplers. These differences are attributable to the “field” components uncertainty, which may include contamination and/or losses during shipping, handling, and sampling, as well as other distortions of the concentration level due to flow and sample volume variations. Uncertainties calculated from the collocation results were found to be generally similar to the uncertainties currently being loaded into EPA’s AQS system, with some exceptions described below.

Final Astro‐2 Calibration of the Hopkins Ultraviolet Telescope

We present the final calibration of the Hopkins Ultraviolet Telescope (HUT) for its flight during the Astro-2 space shuttle mission in 1995 March. Aspects of mission operations and instrument performance that affect data quality are described in detail, as are the data reduction procedures applied to the archived data. The sensitivity calibration is defined by a comparison between synthetic spectra and observations obtained in flight of the hot DA white dwarfs HZ 43, GD 71, GD 153, and G191-B2B; the synthetic spectra were calculated by D. Koester using model parameters derived from fits to ground-based spectra. The resulting flux-calibrated spectra for these stars differ from their respective model predictions by less than 3% at all wavelengths, except at the cores of the Lyman lines where the observed line profiles are shallower than the predicted profiles. As an additional consistency check, the HUT spectrum of BD +75°325 was found to differ from the Hubble Space Telescope Faint Object Spectrograph spectrum by at most 5% in the region of overlap. The wavelength scale and spectral resolution were calibrated by observations of symbiotic stars and of the coronal star α Aur. The spectral resolution was found to vary from 1.8 to roughly 4.5 A, depending on wavelength and on the instrument configuration. The internal consistency of the HUT calibration provides a new and important test of white dwarf model atmospheres, as the predicted stellar spectra are more sensitive to changes in model parameters or to shortcomings in the models themselves at wavelengths shortward of 1100 A than at longer wavelengths. Combining this result with that of the Astro-1 HUT calibration (that a synthetic spectrum of G191-B2B and laboratory flux standards gave results consistent within the laboratory measurement uncertainties) demonstrates that pure hydrogen white dwarfs with effective temperatures of 32,000-61,000 K may be used as primary flux standards down to the Galactic Lyman edge.

Cambridge University Natural Radiocarbon Measurements XV

The dates presented in this paper are concerned with studies of the vegetational history of Western Scotland and are research projects in collaboration with members of staff and research students of the Sub-department of Quaternary Research, Cambridge University. The measurements of radioactivity were carried out between 1975 and 1977 at the University Radiocarbon Research Laboratory at Station Road, Cambridge, using highly purified carbon dioxide as filling gas in proportional counters. The dates are conventional radiocarbon dates calculated using the Libby half-life for the 14C isotope of 5568 years and ad 1950 as the reference year. The associated uncertainties represent one standard deviation and are calculated from a combination of the counting statistics of the samples, standards, and backgrounds together with estimates of population variation, variation of the background rate with changes of barometric pressure, and estimates of other laboratory measurement uncertainties. Thus, the stated uncertainty is considered a fairly reliable estimate of the laboratory uncertainty of the dates. The background samples are prepared from Welsh anthracite and the contemporary standard from NBS oxalic acid. A subsidiary standard is also used which is prepared from the ad 1845 to ad 1855 growth rings of an oak tree which grew near Cambridge, and the activity of this is compared frequently with the NBS standard.