Compact devices for generation of reference trace VOC mixtures: a new concept in assuring quality at chemical and biochemical laboratories

Abstract. Stress-induced volatile organic compound (VOC) emissions from transgenic Grey poplar modified in isoprene emission potential were used for the investigation of photochemical secondary organic aerosol (SOA) formation. In poplar, acute ozone stress induces the emission of a wide array of VOCs dominated by sesquiterpenes and aromatic VOCs. Constitutive light-dependent emission of isoprene ranged between 66 nmol m−2 s−1 in non-transgenic controls (wild type WT) and nearly zero (<0.5 nmol m−2 s−1) in isoprene emission-repressed plants (line RA22), respectively. Nucleation rates of up to 3600 cm−3 s−1 were observed in our experiments. In the presence of isoprene new particle formation was suppressed compared to non-isoprene containing VOC mixtures. Compared to isoprene/monoterpene systems emitted from other plants the suppression of nucleation by isoprene was less effective for the VOC mixture emitted from stressed poplar. This is explained by the observed high efficiency of new particle formation for emissions from stressed poplar. Direct measurements of OH in the reaction chamber revealed that the steady state concentration of OH is lower in the presence of isoprene than in the absence of isoprene, supporting the hypothesis that isoprenes' suppressing effect on nucleation is related to radical chemistry. In order to test whether isoprene contributes to SOA mass formation, fully deuterated isoprene (C5D8) was added to the stress-induced emission profile of an isoprene free poplar mutant. Mass spectral analysis showed that, despite the isoprene-induced suppression of particle formation, fractions of deuterated isoprene were incorporated into the SOA. A fractional mass yield of 2.3% of isoprene was observed. Future emission changes due to land use and climate change may therefore affect both gas phase oxidation capacity and new particle number formation.

Download Full-text

Computation of residual statics using projectors

Geophysics ◽

10.1190/1.1442018 ◽

1985 ◽

Vol 50 (9) ◽

pp. 1502-1504 ◽

Cited By ~ 1

Author(s):

Koraljka Čaklović ◽

Lavoslav Čaklović

Keyword(s):

System Of Equations ◽

Unknown Parameters ◽

Linear System Of Equations ◽

Common Depth Point ◽

Reference Trace ◽

Position Index ◽

Static Corrections ◽

The Common ◽

System 1 ◽

Text Correction

The residual statics problem, as we know, is treated as the solution of one linear system of equations. If we assume that the static corrections are “surface consistent,” then we know that time shifts of each trace can be written as the sum of three terms [Formula: see text] where i = 1, …, [Formula: see text] is the shot position index with [Formula: see text] the number of shot positions, j = 1, …, [Formula: see text] is the receiver position index with [Formula: see text] the number of receiver positions, k = 1, …, [Formula: see text] is the common‐depth‐point (CDP) position index with [Formula: see text] the number of common depth points, [Formula: see text] = correction for ith shot position, [Formula: see text] = correction for jth receiver position, and [Formula: see text] = correction for each trace in the kth CDP gather. For every pair (i, j) we have one equation. We write system (1) in matrix form as [Formula: see text] where [Formula: see text] is the vector of unknown parameters; and [Formula: see text] is the vector which consists of the time shifts obtained by crosscorrelation of each trace in CDP gather with the corresponding reference trace.

Download Full-text

Enhanced local correlation stacking method

Geophysics ◽

10.1190/1.3552687 ◽

2011 ◽

Vol 76 (3) ◽

pp. V33-V45 ◽

Cited By ~ 11

Author(s):

Charlotte Sanchis ◽

Alfred Hanssen

Keyword(s):

Signal To Noise Ratio ◽

Synthetic Data ◽

Weighted Average ◽

Reference Method ◽

Local Correlation ◽

Seismic Image ◽

Reference Trace ◽

Common Technique ◽

Method Yield ◽

The Cost

Stacking is a common technique to improve the signal-to-noise ratio (S/N) and the imaging quality of seismic data. Conventional stacking that averages equally a collection of normal moveout corrected or migrated shot gathers with a common reflection point is not always satisfactory. Instead, we propose a novel time-dependent weighted average stacking method that utilizes local correlation between each individual trace and a chosen reference trace as a measure of weight and a new weight normalization scheme that ensures meaningful amplitudes of the output. Three different reference traces have been proposed. These are based on conventional stacking, S/N estimation, and Kalman filtering. The outputs of the enhanced stacking methods, as well as their reference traces, were compared on both synthetic data and real marine migrated subsalt data. We conclude that both S/N estimation and Kalman reference stacking methods as well as the output of the enhanced stacking method yield consistently better results than conventional stacking. They exhibit cleaner and better defined reflection events and a larger number of reflections. We found that the Kalman reference method produces the best overall seismic image contrast and reveals many more reflected events, but at the cost of a higher noise level and a longer processing time. Thus, enhanced stacking using S/N estimation as reference method is a possible alternative that has the advantages of running faster, but also emphasizes some reflected events under the subsalt structure.

Download Full-text

Data-Centric Performance Improvement Strategies for Few-Shot Classification of Chemical Sensor Data

Engineering Proceedings ◽

10.3390/ecsa-8-11335 ◽

2021 ◽

Vol 10 (1) ◽

pp. 44

Author(s):

Bhargavi Mahesh ◽

Teresa Scholz ◽

Jana Streit ◽

Thorsten Graunke ◽

Sebastian Hettenkofer

Keyword(s):

Error Analysis ◽

Chemical Sensor ◽

Low Cost ◽

Fine Tuning ◽

Sensor Data ◽

Shot Classification ◽

Average Accuracy ◽

Voc Mixtures ◽

Mox Sensors

Metal oxide (MOX) sensors offer a low-cost solution to detect volatile organic compound (VOC) mixtures. However, their operation involves time-consuming heating cycles, leading to a slower data collection and data classification process. This work introduces a few-shot learning approach that promotes rapid classification. In this approach, a model trained on several base classes is fine-tuned to recognize a novel class using a small number (n = 5, 25, 50 and 75) of randomly selected novel class measurements/shots. The used dataset comprises MOX sensor measurements of four different juices (apple, orange, currant and multivitamin) and air, collected over 10-minute phases using a pulse heater signal. While high average accuracy of 82.46 is obtained for five-class classification using 75 shots, the model’s performance depends on the juice type. One-shot validation showed that not all measurements within a phase are representative, necessitating careful shot selection to achieve high classification accuracy. Error analysis revealed contamination of some measurements by the previously measured juice, a characteristic of MOX sensor data that is often overlooked and equivalent to mislabeling. Three strategies are adopted to overcome this: (E1) and (E2) fine-tuning after dropping initial/final measurements and the first half of each phase, respectively, (E3) pretraining with data from the second half of each phase. Results show that each of the strategies performs best for a specific number of shots. E3 results in the highest performance for five-shot learning (accuracy 63.69), whereas E2 yields the best results for 25-/50-shot learning (accuracies 79/87.1) and E1 predicts best for 75-shot learning (accuracy 88.6). Error analysis also showed that, for all strategies, more than 50% of air misclassifications resulted from contamination, but E1 was affected the least. This work demonstrates how strongly data quality can affect prediction performance, especially for few-shot classification methods, and that a data-centric approach can improve the results.

Download Full-text