the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
A python library for computing individual and merged non-CO2 algorithmic climate change functions: CLIMaCCF V1.0
Simone Dietmüller
Sigrun Matthes
Katrin Dahlmann
Hiroshi Yamashita
Abolfazl Simorgh
Manuel Soler
Florian Linke
Benjamin Lührs
Maximilian Mendiguchia Meuser
Christian Weder
Volker Grewe
Feijia Yin
Federica Castino
Abstract. Aviation aims to reduce its climate impact by adopting trajectories, that avoid those regions of the atmosphere where aviation emissions have a large impact. To that end, prototype algorithmic climate change functions can be used, which provide spatially and temporally resolved information on aviation’s climate impact in terms of future near-surface temperature change. These alogorithmic climate change functions can be calculated with meteorological input data obtained from e.g. numerical weather prediction models. We here present an open-source Python Library, an easy to use and flexible tool which efficiently calculates both the individual algorithmic climate change functions of water vapour, nitrogen oxide (NOx) induced ozone and methane, and contrail-cirrus and also the merged non-CO2 algorithmic climate change functions that combine all individual contributions. These merged aCCFs can be only constructed with the technical specification of aircraft/engine parameters, i.e., NOx emission indices and flown distance per kg burnt fuel. These aircraft/engine specific values are provided within CLIMaCCF version V1.0 for a set of aggregated aircraft/engine classes (i.e. regional, single-aisle, wide-body). Moreover, CLIMaCCF allows by a user-friendly configuration setting to choose between a set of different physical climate metrics (i.e. average temperature response for pulse or future scenario emissions over the time horizons of 20, 50 or 100 years). Finally, we demonstrate the abilities of CLIMaCCF by a series of example applications.
- Preprint
(6737 KB) -
Supplement
(864 KB) - BibTeX
- EndNote
Simone Dietmüller et al.
Status: final response (author comments only)
-
RC1: 'Comment on gmd-2022-203', Kieran Tait, 09 Feb 2023
This paper provides an excellent contribution to the field of climate-optimal aircraft routing, with a focus on the spatio-temporal sensitivity of the atmosphere to non-CO2 emission species. The first-of-a-kind ClimaCCF tool is presented, which enables its user to investigate the potential variability in atmospheric response to aircraft emissions, through rapid calculation of algorithmic climate change functions.
The delivery of this manuscript is aptly timed, as the industry becomes increasingly aware of the vast potential to reduce climate impact through operational means, in the interim to fully zero-emissions flight. ClimaCCF is a prime example of a useful tool to communicate complex climate science to the aviation industry, policymakers, and the wider public. In particular, the use of K/kg(fuel) as a metric to distinguish the climate impact of individual flights could be a more useful performance indicator than say, total emissions per flight.
The paper begins with an introductory literature review, which succinctly explains the concept of weather-dependent trajectory planning. It is evident from literature cited, that this project largely builds on the works of FlyATM4E and REACT4. Whilst these research groups have been the driving force behind progress on this topic in recent years, more background information on the fundamental science of non-CO2 climate forcing could be beneficial here. For example, the less experienced reader may appreciate a more comprehensive explanation of (or references to) chemical reactions that lead to NOx-based effects, or the basis of contrail formation and persistence. A clearer explanation of these fundamentals in the beginning of the paper may help to justify decisions made later on regarding parameters used in aCCFs.
The formulation of aCCFs is well presented in section 2. However, the discussion around the use of NOx EI and flown distance per kg fuel is not immediately obvious, based on wording. It took a good few reads to understand that these metrics were introduced purely to change units for the merged aCCFs. Better explanation around this unit conversion for both NOx aCCF and contrail aCCF might help to minimise confusion. See attached pdf for suggestions. In the discussion on climate efficacy conversions, some more detail on how and why RF from different climate forcers results in different levels of warming could help too.
In section 3, the technical details of ClimaCCF are covered in great detail. Figure 1 is a very useful schematic to help understand the process flow of the tool. A reference/link to the exact ECMWF ERA5 dataset used could aid the reader, should they decide to implement the tool themselves.
It is good to see that the science underpinning non-CO2 climate forcing is better presented in section 4, along with the associated example aCCF maps. My research team and I do however, have one major area of concern in the technical implementation of the ozone aCCF: It is stated outright that the photochemical ozone formation does indeed increase with available sunlight. However, ozone aCCF does not take into account irradiance (table 2 states only temperature and geopotential are required to calculate ozone aCCF). How therefore, would these effects be captured in the generated maps? Note, this is less of a point about your findings, and more about drawing attention to the fact that ozone aCCF formulation does not include solar radiation as input. The photolysis reactions pertaining to the formation of ozone are highly sensitive to solar radiation, so more information explaining why the derivation did not identify this sensitivity would be useful.
Sections 5 and 6 round off the paper sufficiently, with discussion around technical implementation and conclusion to highlight key points and findings. One final area that has led to confusion in this paper is the aCCF limitations. Why is it stated that aCCFs are only configured for use in the North Atlantic Flight Corridor region, when all of the maps generated as examples are over mainland Europe? This contradiction between what is stated and what is shown in examples may lead to reader uncertainty.
In general, the findings of the paper were very interesting and engaging, and the tool is an excellent addition to the field. However, technical details need to be addressed such as punctuation, wording and hyphenation, as there were lots of minor technical issues found in the manuscript, hence why presentation quality is given as "fair" in this round of the review. The attached pdf attempts to address the specific areas that may need a second look, so that corrections can be made where deemed necessary.
-
RC2: 'Comment on gmd-2022-203', Anonymous Referee #2, 27 Feb 2023
The manuscript by Dietmüller et al. presents an open-source python library to calculate so-called algorithmic climate change functions for non-CO2 climate effects from aviation. The tool requires information on meteorological conditions as well as certain aircraft parameters like NOx emission index and fuel consumption as input data and calculates the climate impact due to water vapor and NOx emissions as well as contrail formation individually or as combined non-CO2 climate effect. The authors demonstrate the application of CLIMaCCF for a case study in summer 2018 over Europe.
While the tool certainly provides an interesting contribution to the field of climate-optimized flight planning, I find the description given in the manuscript rather difficult to follow. This might be related to the fact that the authors provide a mixture of technical description of the python tool and scientific explanation of the underlying assumptions. Furthermore, the authors refer to several studies that are still in preparation. This is very unfortunate as the reader is left with insufficient information and open questions, e.g., about the calculation of the climate metric conversion factors (sect. 2.4) or the updated set of algorithmic climate change functions, aCCF-V1.1. It might be helpful to restructure the manuscript by starting with the description of the tool, the required input data, the workflow etc., and only then describe the underlying scientific assumptions or simplifications. In particular the description of how the merged aCCFs are calculated (current Sect. 2) needs improvement and clarification. For the authors it is certainly clear what is behind all the parameters, conversion or efficacy factors, and where they come from, but for the inexperienced reader it can be very confusing. Some further detailed comments are given below. Overall, I recommend this paper for publication in GMD after some revisions.
Specific comments:
- L15: which non-CO2 emissions?
- L51/52: I find this sentence a bit confusing. On the one hand the authors talk about climate optimal trajectories, on the other hand they consider only non-CO2 climate impacts. How about additional CO2 emissions that might arise from a re-routing to reduce non-CO2 impacts?
- L90/91: What do you mean by “NOx induced methane”? To me, NOx induced methane sounds like CH4 produced from NOx, but as far as I know NOx emissions from aircraft lead to a reduction of CH4 via NOx induced OH formation, right? So maybe “NOx induced CH4 loss”?
- L94/95: As stated here, the contrail aCCFs are obtained from contrail radiative forcing calculations based on ERA-Interim reanalysis data, but in the example given in the manuscript ERA5 data are used as meteorological input. Do you expect any errors/biases arising from the different meteorological data sets?
- L100: Please provide some more details/examples on the assumptions and simplifications. And how do these affect the results of your tool?
- L106/107: This sentence is not clear to me. How do the different units of the individual aCCFs affect the weighting for different aircraft/engine classifications? Please clarify.
- L147-149: How is this statement related to the values provided in Table 2?
- Table 1: This table provides altitude-resolved average specific NOx emission indices for three different aircraft categories. The values are provided with three decimal places, which implies a high accuracy. However, I would assume that these numbers are associated with some (large?) uncertainties. For example, the wide-body aircraft type seems to include a wide range of different aircrafts. I would be interested to see some uncertainty ranges of these emission indices. Same holds for table 2.
- L182/183: Is the statement on the different emission scenarios a more general comment related to climate metrics or is this directly related to the aircraft emissions? In general, it is not quite clear to me which emission scenarios are meant in Sect. 2.4. Aircraft emissions along a flight track or climate scenarios like the RCPs in general?
- L191: Why is the climate metric P-ATR20 not suited for some questions? And why are F-ATR20/50/100 better suited? Please explain.
- Table3: Why are the conversion factor for H2O aCCF and O3 aCCF identical? Same for CH4 aCCF and PMO aCCF?
- L239/240: “… compatible and tested with the standard of European Centre for Medium-Range Weather Forecasts (ECMWF) data…” What exactly does this mean? Format, naming conventions, meta data? And what is meant by “standard ECMWF data”? Reanalysis? Forecasts? And what would be necessary to use the library with different meteorological data?
- L274: What is meant by “provided default data set”? Does the python library come with a climatology of meteorological data? And if so, where does the default data set come from?
- L289: Why is the PCFA-SAC more accurate and how does is consider aircraft and engine properties? In L282/283 it is written that SAC uses rel. humidity over ice and temperature.
- L368: What is meant by “MET information”? Is MET an abbreviation? If so, please explain.
- 4: Is there any specific reason for using 15 June 2018 as an example?
- A1: Is there any difference in the H2O aCCFs for the different aircraft categories? To me, the plots look identical.
- L408: “.. gets somehow more important…” This formulation is not very scientific and should be rephrased.
- Climate hotspots: I am wondering how meaningful the usage of percentiles as threshold values is? If I understand this approach correctly, it will always identify climate hotspots, no matter how strong the absolute climate impact is, but a re-routing could lead to additional CO2 emissions, so I am wondering how applicable this feature is in practice?
- 5.3: I think this section would benefit from a more quantitative discussion of uncertainties. For example, what is the uncertainty range of the non-CO2 climate effects? Although it is a bit unsatisfying that the authors strongly refer to a paper that is still in preparation. What is the status of Matthes et al, 2022?
- 5.4: Would you expect different results for meteorological input data other than ERA5? How sensitive are the calculated aCCFs to the meteorological data?
- L536/537: What would be necessary to use other meteorological data than ECMWF products in CLIMaCCF? Is it only a coding issue or would the calculation of the aCCF require additional adaptations?
Simone Dietmüller et al.
Model code and software
CLIMaCCF Python Library Simone Dietmüller, Abolfazl Simorgh, Hiroshi Yamashita, Manuel Soler, Sigrun Matthes https://doi.org/10.5281/zenodo.6977272
Simone Dietmüller et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 576 | 206 | 9 | 791 | 62 | 13 | 8 |
- HTML: 576
- PDF: 206
- XML: 9
- Total: 791
- Supplement: 62
- BibTeX: 13
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1