Assessing Novel Lidar Modalities for Maximizing Coverage of a Spaceborne System through the Use of Diode Lasers

Round 1

Reviewer 1 Report

I thank the authors for their work in improving the manuscript to its current state. However, I still take issue with several of the assumptions that are essential to the performance of the solid state vs. diode systems and that have direct bearing on the conclusion that a semiconductor system can do what a solid state mission can do at half the cost. I think the use of the simulator to compare the approaches is of high quality and worthy of publication, but the assumptions and conclusions are not justified properly and give the incorrect impression that a system-level trade study has been performed. If the comments below are addressed (in the text, not simply as a response to the reviewer), I would consider the manuscript ready for publication.

More specific comments are as follows:

 

• Table 1: The Aeolus lidar is not designed to perform altimetry, and was designed with different laser requirements (linewidth, frequency stability, etc.) than contiguous ground coverage. I would suggest removing it from the table and discussion here. Since you are interested in cost I would also include the instrument cost for each instrument, the mission lifetime, the orbit altitude, and the footprint and number of beams. Also, the LITE lidar is not included in Table 1.
• I still take issue with the linear scaling of noise between ICESat II and the hypothetical diode systems. This kind of translation of noise performance across radically different systems (one of which is just a laser at 850 nm) is not justified, even as a conservative estimate (which begs the question, how conservative is it, is it even close to enough to accurate to be useful?). I strongly suggest the authors do the additional work to come up with a complete system design for their 850 nm diode lidar and calculate the expected noise values using data (dark noise, electronics noise, lunar background, etc.) from known subsystems and components. For example, ICESat-II used a Hamamatsu PMT with a counting efficiency of 15%, but what detector would be used for the 850 nm system? The same PMT can not be used at that wavelength with the same performance, and the quantum efficiency and noise performance may be significantly different for an appropriate detector at 850 nm. Similarly, GEDI used a SPAD but at 1064 nm. The Si APD would perform much differently at 850 nm as well, so direct comparison of the laser sources and extrapolation to system performance is not warranted.
• Why is the nightime noise used as a benchmark rather than the daytime or an average of the two? Won’t this give an underestimation of the noise for global mapping or is all mapping planned to be done at night? If this is this case it should be stated as such.
• It is not clear how the lines in Figure 5 were generated. Please reference the appropriate equations used in the text or appendix.
• Line 392 states: “Both of these should give the same ground elevation and canopy height estimates when decomposed into Gaussians and the lowest taken as the ground elevation.” Can you show these results in a figure to validate this claim?
• Figure 7: How was true ground determined? From the ALS data or from airborne radar? It looks in the second frame as if the ALS and true ground are quite different in height.
• The range error “requirement” shifts within the manuscript between less than 4 meters (section 2.1), 3 meters (section 2.3.2), and 5 meters (figure 8). Also, range accuracy and precision are being conflated here. From section 2.1 it sounds like the range precision must be better than 5 meters, but Figure 8 looks like it is measuring range accuracy. Are the requirements for both?
• You cite the ICESat-2 photon rate over ice in your argument for decreasing the PDE to 15% for the solid state laser case. Since your application is for detecting ground under foliage (no mention of global land mapping), the assumption of requiring a multiple element detector does not seem justified.
• What function is being used to fit the curve in Figure 8 and why was that function chosen?
• Line 638 states the sensitives in Figure 12a and 12b are in good agreement overall, but the figures show a 4 to 5x difference in the number of signal photons between the SNR method and the ground return method, which needs to be properly explained if this really is “good agreement”. The ground method also gives an underestimate in the solid state laser case and an overestimate in the pulse train case, which should be discussed. A similar large discrepancy between the SNR method and ground return method results are seen in Figure 13 for the PCL approach.
• Where are the full waveform Gaussian pulse results in the style of Figure 12? The experimental validation is for the PCL case and Figure 8, 9, and 10 refer to that case. Section 4.2.1 was deleted in this version of the manuscript, so there are no results to verify the full waveform column in Table 4, especially the N photons number.
• The authors seem to be applying different standards to the technology readiness of the solid state and pulse train systems in assessing their viability and potential cost savings. For example, they state that Yb:YAG systems with 11% efficiency have been demonstrated, but are not considered because they have not yet flown, yet no diode laser system of this kind has flown, but the performance of the pulse train system is assumed to equal that of cutting-edge research-grade tapered diode lasers. To perform a fair comparison of the techniques, a uniform standard should be applied to both technologies. Either choose only flown performance or best-in-class R&D performance.
• The discussion of possible swath widths (and a single-spacecraft mission concept) does not take into account the detector or signal processing elements required to actually implement such a system. For example, the 10,150 m swatch width for a pulse train system would require 339 detector elements, each with their own timing systems for waveform capture. How would this affect the instrument/satellite power and cost to implement such a system? There should at least be some discussion of the system-level trades and limitations that would go on in designing a real system such as this.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Manuscript entitled “Assessing Novel Lidar Modalities for Maximizing Coverage of a Spaceborne System Through the Use of Diode Lasers” does an excellent job by demonstrating the significance of diode lasers in spaceborne instruments. This is an very important issue as existing spaceborne airborne laser scanning is limited to sparse sampling missions due to energy requirements and high cost. Alternatively, technological advancement in diode lasers may benefit the ALS to use in wide range of applications.

 

Regarding the content, honestly I was pleased to see how the article organized. It is well written with very clear references, detailed analysis and comparison. And would be very appreciable, if abbreviations provided for each technical short word used in the manuscript (ex. GEDI, ICESat, so on). Other than this, i believe this is a complete manuscript and there are no any modification required as in the current format.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The manuscript has been improved and the conclusions clarified in this draft. I consider this version suitable for publication.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The paper investigate the potential for using diode lasers, with their higher efficiencies, as an alternative. The contribution of the work is quantified, followed by the  efficiencies of about 25% and are much smaller and lighter than solid-state lasers.The diode laser can be operated in pulse train or pulse compressed lidar (PCL) mode from space, using a photon counting detector.I generally enjoyed their work and it gives value to the literature as long as the following are satisfied. Generally, I believe this paper is well written.

Reviewer 2 Report

This authors have performed a model comparison of two diode based laser systems to the current orbital instruments GEDI and ICESat-2, with the goal of determining if a diode-based system may be more efficient for measurements through foliage. The manuscript makes extensive use of a previously published GEDI lidar simulator, though the only comparison of the simulator results with experiment (the PCL bench experiments) do not show very good agreement. The concept of designing a diode-based system to perform these measurements is worth investigating, but the methods used by the authors are too vague and imprecise to extract much useful data from the overly long manuscript. Some general comments are:

• This manuscript has elements of merit buried within it, but they only make up mostly the latter portion of the overall text, and the length and wandering nature of the paper makes it hard to keep the through-line. Importantly, the overall conclusions as written are not of much use as so many under and overestimates have been made throughout that there are no clear targets for instrument designers to target in their efforts.
• The title is not accurate as written. There is no evidence that they have determined the "optimum" lidar modality, and it doesn't read as if maximizing coverage is the primary goal, rather, the effort seems to be on finding the minimum average and peak laser powers required.
• There is a distinct lack of discussion about ranging accuracy, which is just as important as getting a ground return through foliage. This stems more broadly from a lack of clear measurement requirements that all the compared systems must meet that should be laid out right after the introduction and justified with science tracibility.
• The introduction does not adequately set the stage for the following sections and is far too light on discussions on the key design drivers for ICESat-2 and GEDI, which the authors intend to offer an alternative design to. The introduction focuses on the idea of complete coverage using constellations of lidars and system cost, but never returns to this idea numerically or otherwise. Similarly, the discussion of coverage and swath width are presented at the very beginning of section 2 but do not factor into the rest of the work at all. Only the metric of success, Psi, is discussed throughout. Really the paper is all about the minimum detected energy per measurement. Poorly justified and broad assumptions about Q and Le make it difficult to have confidence in the end values of Psi.
• The paper relies heavily on the lidar simulator from Hancock et al., but in several places the simulator is not shown to agree well with experimental results presented. The authors generally just trust the simulator and describe the results as “pessimistic” or “optimistic” rather than work to determine where the simulator is in error (a much more useful task).
• Overall, there is a disconnect between the parameters of the laser modalities being tested and what can be feasibly built for flight in the near term. I would encourage the authors to collaborate with instrument scientists working on diode laser systems to get a better sense of what is feasible and start there with their performance models. A better way to look at this problem may be to try to design a PCL or pulse train system model with reasonably high TRL subsystems (laser, optics, detector) that are referenced and justified properly. From here you can create a noise model from scratch based on subsystem performance that is more accurate than just scaling ICESat-2 nighttime values linearly. The system will likely not be better than ICESat-2 or GEDI but at least you can say something about how a current system may fare and where the points of improvement lie.
• The use of simple linear scaling of noise sources is not appropriate, especially as the laser wavelengths, detector types, and laser sources are all varied from the baseline ICESat-2 design. This gives the impression of precision for things like the minimum detectable energy that are not validated.
• The text of all the figures is generally too small to comfortably read without zooming way in and would be illegible if printed. The captions also should provide enough information for the figure to be completely understood without searching within the main text for crucial details. Ensure all figures have labelled axes and are frames are not obscuring one another

Please see the attached document for  specific comments with line numbers referenced.

Reviewer 3 Report

This is well written and well thought manuscript. Although not from the same area of scientific interest, I learnt from the manuscript for my own laser-based work.

I have some minor critics, which I would like to recommend the authors to improve: the figures are lacking very often axis titles and/or units. Right from Fig. 1, it is open to the readers creativity what is depicted on the x-axis. On this figure also the colous of the 15 and 30 ns pulse curves are hard to distinguish. Figs. 2, 3, 4 don't have an y-axis title nor a unit for it. Fig. 5 comes without a single unit to all its layers. On Fig. 6 the x-axes labels are overlapping between different layers. Fig. 9 has some y-axes called 'Photons', what is not even a quantity and if this figure means photon number or counts, I wonder what a photon number below 1 really means. Fig. 9 is generally a bit overcrowded with data and curves.  The y-axes Fig. 11, 14, 15, 16 read beam sensitivity but come without a unit. Sensitivities normaly do have units. If this is supposed to be a relative sensitivity, it should read like this.   

Very often units are given in '[]'-brackets. The authors may have a look how units are supposed to be specified. To my knowledge is the bracket type '[]' not a recommended way.