the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 May 2025
In-flight emission measurements with an autonomous payload behind a turboprop aircraft
Abstract. This paper reports on successfully testing a new, autonomously operating measurement system on a Grob G 520 Egrett aircraft for in-flight aerosol and trace gas measurements of engine exhaust. A suite of in-house-built and commercially available instruments was selected, modified, and adapted to the unpressurized compartment of the Egrett to operate over a wide range of ambient temperatures and pressure levels. We performed first in-flight emission measurements at cruise altitudes behind a twin-turboprop aircraft, the Piper Cheyenne, powered by Garrett/Honeywell TPE 331-14 engines, over Texas in April 2022.The instrumentation and inlets on the Egrett were designed to measure non-volatile particulate matter (nvPM), total particulate matter (tPM), nitrogen oxides (NO and NO2), water vapor (H2O), carbon dioxide (CO2), and contrail ice particles. All instruments were operated in relevant plume conditions at cruise altitudes between 7.6 and 10.4 km (FL250 and FL340) at distances ranging from 100 to 1200 m between the two aircraft. The instruments proved to have high reliability, a large dynamic range, and sufficient accuracy, which is adequate for measuring the emissions of the turboprop engine.
We derived the emission indices (EI) for tPM, nvPM, and NOx at cruise. The particulate emission indices range from 9.6 to 16.2 ×1014 kg−1 (particles per kg fuel burned) for EItPM and from 8.1 to 12.4 ×1014 kg−1 for EInvPM (medians). For NOx we find rather low EINOx between 7.3 and 7.7 g kg−1 for EINOx (medians). Furthermore, aerosol size distributions have been measured in the exhaust plume. The analysis of the size-resolved emission index indicates a log-normal distribution with geometric mean and standard deviation at Dg = 34.7 ± 1.9 nm. This geometric diameter value is in the range of jet engine soot emissions previously measured in flight. The measurements help to constrain the climate impact of current turboprop engines and provide a benchmark for future alternative H2 propulsion systems such as fuel cells and direct combustion engines.
Competing interests: CR and JC are employed by Airbus Operations. RV and AV are employed by AV Experts LLC. All the other authors declare that they have no conflict of interest.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(21444 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2026', Anonymous Referee #1, 11 Jun 2025
-
AC1: 'Reply on RC1', Gregor Neumann, 26 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2026/egusphere-2025-2026-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Gregor Neumann, 26 Sep 2025
-
RC2: 'Comment on egusphere-2025-2026', Anonymous Referee #2, 05 Aug 2025
The paper describes a new in-flight measurement capability for measuring a range gaseous and particle emissions from aircraft engines. The work describes the first application of the method measuring behind a turboprop. Data from turboprops, both in-flight and on the ground, are rare and this provides an extremely useful dataset to the community. The application of two inlets, allowing for simultaneous in-plume and background, is a really nice feature.
The paper is very well written, with particular attention paid to the detail. The uncertainty in the measurements is well treated. I only have a few minor comments, which are listed below.
In the abstract, define cruise altitude (Line 5)
Line 14: do you mean lower than expected NOx based on predicted? Please clarify (based on previous, based on ground-based predictions?). Could it be caused by using H2O instead of CO2 for the calculation (see below)?
Line 14, tPM, nvPM or vPM size distributions?
Line 29, delete comma after Both
Line 41, do you have a reference to a roadmap or similar to the potential for these new technologies?
Figure 2 is referenced before figure 1 (line 87).
There needs to be a bit more detail on the operational details of the Egrett (operating altitudes, range, science speed range)
Figure 5b. The model is outside of the error bars of the data exactly at pressures where most of the data is collected. There either needs to be more points added to constrain the fit better or this needs to be incorporated into the error analysis.
Figure 5d, It is not clear what the data is. The Y axis is labelled as aMCPC but the figure legend is inlet system losses. Can this be explained more clearly and which curve or curves are used in the loss correction section?
Line 267 – why is there a range of sheath flows? Have you verified this is not changing during a scan? Has this been incorporated into the error analysis (changing the sheath to aerosol ratio changing the resolution of the DMA etc)?
Line 464 – where does the value of 10% undercounting and subsequent correction come from? That needs clarification.
The EINOx using the water vapor is an interesting approach. I can see not having CO2 co-located with the Nox at short distances might be an issue as one inlet may be in the plume and the other not. What I would like to see is the EInumber calculated with CO2 and H2O(CR2) from the box A inlet as these should give the same value and give confidence in the EINOx approach based on the author's claim that the WARAN and CR2 agree well.
Section 2.1.8, 4.4 and figure 9 – the paper goes into great detail on the uncertainties in the measurement system, but I do not see that same detail for the mSEMS or the OPC. For the mSEMS, the extremely low charging probabilities make quantification challenging. 100 particles at 10nm in dN/dlogDp space over a 5 second scan, corrected for charging efficiency, is an incredibly small number of particles getting to the MCPC detector. Is the variability in the data as shown by the red shaded area in figure 9 really larger than the uncertainty associated with the instrument and conditions (short scan time (smearing), low numbers, charging probabilities, possible changing DMA resolution)? Are the smallest sizes in the distribution a true representation of the PSD?
Purely out of curiosity, given the relatively simple equations being used, would Monte Carlo simulations be a simpler and more accurate method of calculating the uncertainty rather than the full error equation?
Citation: https://doi.org/10.5194/egusphere-2025-2026-RC2 -
AC4: 'Reply on RC2', Gregor Neumann, 26 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2026/egusphere-2025-2026-AC4-supplement.pdf
-
AC4: 'Reply on RC2', Gregor Neumann, 26 Sep 2025
-
RC3: 'Comment on egusphere-2025-2026', Anonymous Referee #3, 20 Aug 2025
The manuscript describes the instruments comprising an autonomous payload for measurement of CO2, NOx, H2O and particles and their initial deployment on a Grob Egrett aircraft to sample the exhaust emissions from a light turboprop aircraft at cruise altitude. The manuscript is well organized and the descriptions and analysis are clear and fairly comprehensive. The topic is certainly relevant for AMT and I recommend it for publication with only minor comments and suggestions for the authors.
Minor comments:
L2: perhaps “on the successful first deployment of”
L3: could clarify that you are measuring the exhaust of other aircraft
L3: perhaps “custom-built and commercially”
L5: “temperatures and pressures”; “performed these first”
L6: suggest “a Piper Cheyenne, a twin-turboprop aircraft powered by…”
L11: suggest omitting “, which is adequate”
L74: “non-CO2 effects from aircraft emissions”
L75: “of the instrument payload for a chase aircraft”; size distribution not included here?
L81: Could omit paragraph, or be a little more explicit—“Further” seems vague
L88: “near-field exhaust plume”
L96: omit “specifically” and maybe “accommodate”
Fig 1 caption: “Piper Cheyenne (400LS, registration 30 N92EV)”
L130: the heated section evaporates the volatile material, the subsequent cooled section is to lower the temperature prior to introduction into the CPC, right?
L136: omit the last sentence
Fig 3: “puring” “”purging”; the pumps associated with the aMCPCs are for the saturator flow, not a sheath flow, correct? The mSEMS does have a sheath flow—pump not shown?
L166: “as is shown in Fig. 6.”
Fig 5 (and subsequent uncertainty discussion): it would be nice to have a panel that shows the combined uncertainty of the various factors that are shown separately and a discussion of the overall magnitude to conclude section 2.1
L212: “or do not grow large enough”
L223: “deposition” would be a better word than “sedimentation”, or you could just say “diffusion to the tubing walls”
L226: why do ground-based measurements necessarily require longer inlet lines and residence times?
L230: is the heating of the sampling line “to avoid significant losses of small particles on the tube walls” mechanism thermophoresis? Or are you preventing ice build-up? How warm? For what length is there a thermal gradient?
L246: inner diameter?
L259: “soot soot”
L362: “referred to as “particle” speed because that is the speed CAPS observes particles to travel? Otherwise “True Air Speed” is the more recognized parameter
L375: “in situ” is not hyphenated
L382: “as a dilution”
L401: “near-field”; “measurements, as inside contrails and clouds condensation makes water vapor non-conservative.”
L406: Schumann ref in parentheses? Sig figs on molecular seem excessive—actually could omit the number altogether.
L414: “vertical profile”
L421: “example”; “emissions of”
Fig 7: time series of Nnv / Nt would be interesting to see; “near-field” in caption
L437: “on the order of”
L450: Clause including “slightly aged about one-minute-old” is awkward
L461: First sentence is unnecessary
L465: “low pressure counting”
L466: not sure what is meant by “corresponding”
L493: compare to what ground and in-flight measurements? “previous” of …
L543: what is “jet-phase”?
L547: “data are archived”
Citation: https://doi.org/10.5194/egusphere-2025-2026-RC3 -
AC2: 'Reply on RC3', Gregor Neumann, 26 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2026/egusphere-2025-2026-AC2-supplement.pdf
-
AC3: 'Reply on RC3', Gregor Neumann, 26 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2026/egusphere-2025-2026-AC3-supplement.pdf
-
AC2: 'Reply on RC3', Gregor Neumann, 26 Sep 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 688 | 113 | 17 | 818 | 19 | 36 |
- HTML: 688
- PDF: 113
- XML: 17
- Total: 818
- BibTeX: 19
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Summary: The impact is well-motivated, showing the importance of studying short range flights with turboprop aircraft. Significance of the impact, difficulty of the measurement, and lack of comparable datasets highly motivate the scientific significance. The method is sound and calibrations and subsequent analyses are thorough, including variance and error propagation. Some explanation as to why the aerosol instruments need to be pressurized would be preferable, since I was confused by the explanation of sampling stability. I am suspicious of the particle sizing and concentration accuracy at the small sizes ~10nm, due to the cut size of the particle counters (exacerbated by diffusion losses) leading to large correction factors, thus the mode size may be slightly overestimated; however, results compare well to previous airborne measurements with similar instrumentation from Moore et al., 2017. The author did a very good job characterizing instrument response to sample and environmental factors, and uncertainty. There are a couple of typos.
L15-17: Are the size distributions presented from the total or nonvolatile aerosol? If total, suggest removing “soot” and generalizing as jet engine emissions, since the soot implies non-volatile particulates.
L25-26: Incomplete sentence. Suggest combining with previous sentence.
L97: L228 is the first time bringing up isokinetic, but it would be helpful to mention at the introduction of the aerosol inlet.
L122-123: Can you expand on what “ensuring stable sampling conditions” means? Why would the instruments need to be in a pressurized vessel? It introduces a higher deltaP and increases the potential for dilution/leak into the sample.
L206-208 & F5b: The curve fit deviates from the measured counting efficiency right around your critical operating environment. The operating environment is barely captured in your data points. Higher resolution in the region that you measure (more points between 375-250 hPa) would result in more precise correction.
L223: Is the 90-degree tube bend sufficiently large enough radius to be negligible for inertial impaction of large particles? Valuable to mention if negligible here when describing other loss mechanisms.
L259: Typo. soot soot.
L267: Why the range in sheath flow? Is the range from intentional changes, e.g., compensating for pressure to achieve the same size range in a scan, or fluctuation due to environment? What is the corresponding sample flow?
L270: Typo on mSEMS. “Is able to operate at 5 s scan time”, does that mean you did operate a 5 s scan? What was the lag time from the sample out of the DMA to the aMCPC? For a 5-second scan, the smearing may be significant. You don’t mention operation scan times until L509, and it’s worth mentioning how it was operated in section 2.1.8. 17s scan while in a highly variable plume seems too slow for samples shown in F7 without a large lag chamber. Were scans averaged to suppress the noise, and if so, how many scans are used for averaging?
L289-290: Is it supposed to read “sizing” instead of “size”? Why are instrument sizing and flow calibration major sources of uncertainty? Are you talking about the physical size due to unknown refractive indices?
F9, F5c, F5d: The combined effects of the aMCPC size cut, counting efficiency with pressure, and diffusion losses, hurts the confidence in the size distributions below 20 nm. I expect the entire left-side falling edge of the curve in Figure 9 in plume would have increasing error bars associated with it, which may provide context/caution in interpreting the mode size from the fit in F9b. The aMCPC may not be the best choice for engine emission characterization since its cut size is near to the exhaust particle size range. There may be a significant number of sub-10 nm particles missing from the tPM when calculating the EI. Perhaps the EI should be specified as the EI_tPM>10nm at the top of the document. This may be less of a concern for this generation of engine, but consider ultra fine CPCs when testing future generation engines that combust more efficiently that they may have a smaller mode size where the aMCPC will completely misinform/bias the peak.