Comparative Analysis of Compact Portable and Indoor Rainfall Simulators

Falcão, Raquel N. R.; Krása, Josef; Neumann, Martin; Kubát, Jan-František; Gall, Corinna; Seitz, Steffen

doi:10.5194/egusphere-2025-5908

Preprints

https://doi.org/10.5194/egusphere-2025-5908

Preprints

04 Jan 2026

| 04 Jan 2026

Status: this preprint is open for discussion and under review for Geoscientific Instrumentation, Methods and Data Systems (GI).

Comparative Analysis of Compact Portable and Indoor Rainfall Simulators

Raquel N. R. Falcão, Josef Krása, Martin Neumann, Jan-František Kubát, Corinna Gall, and Steffen Seitz

Abstract. Rainfall simulators have been widely used and are indispensable in soil hydrology and erosion research. Although rainfall simulators can be used to address various research questions, there is no standardized methodology; they differ in design and rainfall characteristics. The present study aims to describe the design and testing of five rainfall simulator setups that vary notably in weight, volume, and transportability. Additionally, this article seeks to clarify procedural aspects involved in conducting a rainfall simulation in the field. The following parameters are used to compare the simulators: drop size distribution and terminal velocity, uniformity of the spatial distribution of raindrops over the sprinkled surface area (Christiansen uniformity coefficient; CU), kinetic energy (KE), and rainfall intensity. The Thies laser disdrometer and Tübingen Splash Cups (T-cups) were used to measure raindrop's KE to identify similarities and differences in their rainfall characteristics.

The rainfall simulator setups produce rainfall intensities ranging from 28 to 95 mm h^-1, with CU values ranging from 60.5% to 75.8%. More than 90% of measured drops were slower than 3.8 m s^-1 for all simulations. The maximum number of drops was below 0.5 mm class, generally smaller than that observed in natural rain, and all at 1.4–1.8 m s^-1 velocity. We found that kinetic energy (KE) measured with T-cups agreed with values calculated with the Thies disdrometer, confirming its relevance in rainfall studies. Indoor simulator setups produced the highest KEs, whereas the portable systems showed considerably lower values.

This study emphasizes the importance of accurately characterizing rainfall parameters before soil erosion measurements. Rain simulators are then a powerful tool in erosion research. The presented methodologies and insights provide means for improved assessment of soil erosion risks, particularly regarding their practicality in remote areas.

Received: 29 Nov 2025 – Discussion started: 04 Jan 2026

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.

Download & links

Raquel N. R. Falcão, Josef Krása, Martin Neumann, Jan-František Kubát, Corinna Gall, and Steffen Seitz

Status: open (until 09 Feb 2026)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2025-5908', Anonymous Referee #1, 05 Jan 2026 reply
Dear Editor
Thank you for giving me the chance to review the attached manuscript, hinging on the applicability of different rainulators to simulate rainfall under various circumstances. The work is potentially worth it; however, it suffers from serious issues listed below and annotated in the reviewed manuscript.

What specific problems does this study address? Each simulator operates under different conditions, which leads to discrepancies in the results. How can you justify or account for these discrepancies? No reference is made to natural rainfall conditions or the method used for measuring raindrop characteristics.

The abstract must clearly present the research significance, materials and methods, results, conclusions, and recommendations.

The text mentions five types of simulators... The type of simulator must be specified.

On what basis were these selected?

Most of the information on the simulator designs is incomplete. The type of simulator, nozzle, and other details must be specified, and the results should be accurately presented for each simulator.

The main reason for this study is not specified in the introduction.

Some paragraphs are unnecessarily long in structure.

Failure to present the main hypothesis.

Repetition of objectives!

The grammar and language of the manuscript should be improved.

New references should be used.

The introduction lacks a direct and quantitative link established between these technical differences and the final erosion outcomes (such as sediment yield or runoff). This creates a gap between "the characteristics of the simulated rainfall" and "its ultimate purpose (studying erosion)".

To what extent are the findings from these devices generalizable to the broader global community of rainfall simulators?

Can general principles be derived from the results?

Although "being the first" is an innovative point, the introduction does not explain why this work is important.

Why was this cup used and not other conventional methods?

Why is comparing these three specific devices important?

Equipment design and installation are missed.

To what extent can environmental factors such as air humidity, water pressure, etc., be influential?

It should be noted that these experiments were conducted in different regions, each with its own climatic parameters, which could affect the results.

When the height is variable, the results will certainly change as well.

In conditions outside of a rainfall simulator, how do you account for the effect of wind on raindrop characteristics?

The measurement of raindrop characteristics was not conducted under uniform conditions.6: Has the plot effect been considered in this study?7: Calculate other properties of raindrops, including area, perimeter, angle, and external energy.8: This section remains unclear.

Furthermore, there is no information on natural rainfall conditions.

The results are not very concise and clearly stated. It is necessary to provide detailed results of the characteristics of raindrops in each of the simulators.

How do you compare the results with different rainfall intensities? The intensity of normal rainfall is still unknown.

The value of this coefficient is low. How do you justify it?

in discussion section, what specific aspects of nozzle design might lead to smaller droplet production?

How can these findings be used to improve the accuracy of future precipitation simulations?

How do these deviations compare with similar results from other studies? Could the specific circumstances of this study be the main reason for these differences?

Given the CU values, how can a more accurate or improved criterion be reached for evaluating the precipitation distribution in simulations?

Could such differences have major impacts on soil erosion simulations and hydrological models? How can these differences be accounted for in future modeling?

Are there more accurate methods for measuring kinetic energy of rain that could improve these analyses? What factors cause T-cups measurements to be inaccurate in low kinetic energy rain?

Explain the relationship between kinetic energy of rain and soil erosion.

Given the high costs of conducting replicate experiments, what suggestions do you have for improving the accuracy and efficiency of these methods in the context of cost-oriented research?

How can these effects be more accurately incorporated into soil erosion models to achieve better predictions of soil erosion?

What solutions can be adopted to increase the scalability of these simulators without reducing accuracy and measurement capabilities?

Can new technologies be used to develop these systems?

Study limitations and suggestions and future research should be included in the conclusion.

Considering the above-mentioned comments, flaws, and the potential of the manuscript, it requires substantial revision (border rejection).

Reply
Citation: https://doi.org/10.5194/egusphere-2025-5908-RC1
RC2: 'Comment on egusphere-2025-5908', David Dunkerley, 09 Jan 2026 reply

This paper describes several ‘rainfall simulation’ devices from Czechia and Germany. The rationale for the paper was not clear to me, given that across the world, such devices vary enormously in scale, design, and operation. The three basic devices described are scarcely comparable in scale or sophistication to many in use globally, including the enormous Tsukuba facility in Japan, which can work with plots of up to ~ 3000 m², and which can work across a far wider range of rainfall rates than the devices described by Falcão et al. I recall working with Athol Abrahams (1991) on an 18 m x 35 m plot (area 630 m²) in Arizona.
            Such large plots – surely designed at their considerable size for good reasons – raise a fundamental issue: why do the authors consider that plots of just a couple of square metres (or even < 0.2 m², as in the case of the Tübingen portable device) can represent a field, or a hillslope? And how could this be the case, given that the plot designs mentioned all have steel walls that exclude runon water that would under natural conditions arise from upslope, where it would also be raining? Because of this, the depth of water on the plots would be unnaturally low, and its erosivity therefore also too low. How then do the authors scale-up their results to what would be expected under deeper, faster, natural surface runoff?
            The second issue is a related one. Given that the devices reviewed by Falcão et al. use different rainfall rates, drop sizes and fall speeds (and some are in addition pulsed or intermittent in their application of water, and they all use different plot sizes and shapes), which result would be most representative, if all were used on adjoining plots in a single field test site? In other words, does the use or one or another of these devices affect the results obtained? If so, which is more believable? Clearly, this suggests that if such devices were to be used, say, in a serious attempt to understand erosion processes for a land manager, the results would need to be validated by comparison with observations made under natural rain (for which all parameters, drop size, KE, fall speed, etc. are ‘correct’), and which remains the ‘gold standard’ for studying hillslope processes. A good example is the work of Liang et al. (2023), who employed field plots exposed to natural rainfall in their study in China.
            Clearly the intensities and spray systems of the several devices described in the paper were selected for what might be termed ‘hardware’ reasons (e.g., to use components that were readily available from agricultural sprayer manufacturers, available at low cost, etc.) and not for reasons relating to soil hydraulics or erosion.
            The authors seem to consider (e.g. lines 329-330) that simply being reproduceable (at least for each device, although they are clearly very different) is an asset. For most applications, I disagree. Having a ‘test’ that is reproduceable is not sufficient. Consider as an analogy crash-testing motor cars, all at a speed that is too low (like the runoff speed on small, bounded plots) – say 10 km h^-1. Would you buy a car whose safety was advertised on the basis of such reproduceable test results? (And likewise, if you were trying to understand and manage soil erosion, would you trust small, bounded plot results simply because they were reproduceable?). In both cases, I would suggest that the answer is clearly ‘NO’. In both cases, speed (of the car or of the surface runoff) MUST be realistic or else the results will not be relevant to the real world.
            What the authors must do, therefore, is to identify and describe those problems that they consider can be validly explored using the devices described, giving justifications for their views. Given that the reviewed devices do not ‘simulate’ natural rain at all well, the authors need also to say how they have validated their results. (In the same way, a lab crash-testing cars at 10 km h^-1 would surely be expected to justify their odd procedure, or else people would completely disregard their test results). There has been remarkably little attempt to validate rainfall ‘simulation’ work, and the limited available results are not particularly encouraging (Dunkerley 2021 a).
            I should add a few additional items which cast doubt on many kinds of study likely to be undertaken with the devices described by Falcão et al.:
1) The areal rate at which drops strike the soil surface (thereby breaking aggregates, building raindrop impact crusts and seals, driving air into the soil pores, and so on) is critical to get right (Dunkerley 2021 b). It cannot be correct if the drop-size distribution is not correct.
2) As the authors mention, intensity during natural rainfall is NEVER constant, and we know that the shape of the intensity profile exerts a major influence on runoff ratio, time to peak, and so on (Dunkerley 2021 c). This cannot be neglected, or else how are we to argue that the rainfall ‘simulation’ results – however reproduceable – actually relate to the natural world and its mechanics? If we don’t get the intensity profile right – say, most intense early in a storm as with convective storms, or most intense late in the event – then we cannot even describe soil hydraulics or erosion as it would occur in different seasons, when the intensity profile is likely to be different.
3) The DURATION (as well as the intensity profile) of a rainfall ‘simulation’ should match the natural storm characteristics of the study area, including any intra-event breaks in rainfall (intermittency). Falcão et al. fail even to specify the test durations that are achievable with the several devices mentioned. By way of illustration, consider just one example of many more considered or refined studies: Frauenfeld and Truman (2004) – more than 20 years ago! – worked on runoff and interrill erosion on soils from the coastal plain of Georgia (USA). They compared constant-intensity and variable-intensity experiments. They even designed the variable-intensity experiment on the basis of detailed rainfall records from their field area – it was not made up arbitrarily! Total erosion in the variable-intensity events was up to 3 x greater than for the fixed intensity, even for the same duration. More studies of this kind were reviewed in Dunkerley (2012). We have apparently regressed in our methods since then, and lost sight of the importance of duration and intensity profile.
Summary:
The devices described by Falcão et al. do not seem to be able to simulate rainfall appropriately, even as regards drop size or fall speed, let alone intensity profile or drop arrival rate. Other factors mentioned in my review appear to be quite beyond the capacity of the devices. I cannot therefore see this as a paper that would be of use to researchers unfamiliar with the methods discussed, since they are not even described in full detail (e.g. lacking any mention of rain duration). A more thorough review of the wide range of devices currently available would make a better and more useful contribution.

My recommendation to the Editor is that the manuscript as it stands should be rejected.
It is worth commenting that there was an evident lack of care in proofreading, since several odd items appear in the text as submitted:
Line 67: “Click or tap here to enter text.”
Lines 69-70: “Click or tap here to enter text.”
Line 243: (Error! Reference source not found.).
Items of this kind should really be removed prior to manuscript submission.
David Dunkerley

Faculty of Science

Monash University
References
Abrahams, A. D., & Parsons, A. J. (1991). Resistance to Overland Flow on Desert Pavement and Its Implications for Sediment Transport Modeling. Water Resources Research, 27(8), 1827-1836. doi:https://doi.org/10.1029/91WR01010
Dunkerley, D. (2012). Effects of rainfall intensity fluctuations on infiltration and runoff: rainfall simulation on dryland soils, Fowlers Gap, Australia. Hydrological Processes, 26(15), 2211-2224. doi:10.1002/hyp.8317
Dunkerley, D. (2021 a). The case for increased validation of rainfall simulation as a tool for researching runoff, soil erosion, and related processes. CATENA, 202, 105283. doi:https://doi.org/10.1016/j.catena.2021.105283
Dunkerley, D. (2021 b). Rainfall drop arrival rate at the ground: A potentially informative parameter in the experimental study of infiltration, soil erosion, and related land surface processes. CATENA, 206, 105552. doi:https://doi.org/10.1016/j.catena.2021.105552
Dunkerley, D. (2021). The importance of incorporating rain intensity profiles in rainfall simulation studies of infiltration, runoff production, soil erosion, and related landsurface processes. Journal of Hydrology, 603, 126834. doi:https://doi.org/10.1016/j.jhydrol.2021.126834
Frauenfeld, B., & Truman, C. (2004). VARIABLE RAINFALL INTENSITY EFFECTS ON RUNOFF AND INTERRILL EROSION FROM TWO COASTAL PLAIN ULTISOLS IN GEORGIA. Soil sci, 169(2), 143-154. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=ovftg&NEWS=N&AN=00010694-200402000-00007
Liang, Y., Gao, G., Liu, J., Dunkerley, D., & Fu, B. (2023). Runoff and soil loss responses of restoration vegetation under natural rainfall patterns in the Loess Plateau of China: The role of rainfall intensity fluctuation. CATENA, 225, 107013. doi:https://doi.org/10.1016/j.catena.2023.107013

Reply

Citation: https://doi.org/10.5194/egusphere-2025-5908-RC2
RC3: 'Comment on egusphere-2025-5908', Jesús Rodrigo-Comino, 23 Jan 2026 reply

As the third reviewer, and having read the previous reviews from my colleagues, I will not repeat the comments made by Prof. Dunkerley, with most of which I agree. I think these points need to be better justified in the manuscript in order for the paper to be published with solid scientific support. In addition, having worked with some of the authors in the past, and with some of these or similar equipment, I trust the effort made by the authors when carrying out each rainfall simulation, especially those using large-size simulators. However, I believe that in its current form the paper cannot be published in a high-impact specialized journal and requires major revisions.
The title is too broad: what type of soils, rainfall intensity, and geomorphological conditions are being addressed?
Regarding the abstract, the very first sentence is too strong. Rainfall simulators are useful, but not indispensable.
I do believe there are standardized methods, but not globally for all simulators. The authors state that they aim to clarify aspects of field work with rainfall simulators, but they do not specify what exactly is still unclear.
Several statements about disdrometers and calibration procedures refer to well-known issues and do not seem novel.
I miss references to studies where these simulators have been applied in real conditions, in order to better understand how they perform and what results they produce.
I do not find it appropriate to include figures taken far from the simulator, as they do not show its components, the fieldwork, or schematic diagrams with its characteristics. I would even consider including videos as supplementary material to validate the work and address some of the issues rightly pointed out by other reviewers. For example, in Figure 3, neither the simulator nor the soil surface can be seen.
Not describing the soil type and its initial conditions is, in my opinion, a major flaw, as it prevents a full understanding of the experimental setup.
The statistical analysis could be expanded slightly, including information on the libraries used and even the code to reproduce the figures, since the paper is presented as a study that should be reproducible by other researchers.
I do not fully understand why the figures are not properly numbered, for example Figure 8 (a, b, c).
There are many citation typos, likely related to Zotero or Mendeley, including missing references.
Personally, I do not like the black-and-white figures, particularly for the heatmaps, and I wonder whether this could be improved.
The discussion relies heavily on old references, and I do not clearly see the development of what is stated in the abstract and objectives, namely guiding authors towards standardized work and protocols. The objectives may need to be reformulated.
I have included additional comments and annotations in the PDF, specifically in the sections where I believe changes are necessary.

Reply

Citation: https://doi.org/10.5194/egusphere-2025-5908-RC3

Raquel N. R. Falcão, Josef Krása, Martin Neumann, Jan-František Kubát, Corinna Gall, and Steffen Seitz

Viewed

Total article views: 241 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
157	68	16	241	11	9

HTML: 157
PDF: 68
XML: 16
Total: 241
BibTeX: 11
EndNote: 9

Views and downloads (calculated since 04 Jan 2026)

Month	HTML	PDF	XML	Total
Jan 2026	157	68	16	241

Cumulative views and downloads (calculated since 04 Jan 2026)

Month	HTML	PDF	XML	Total
Jan 2026	157	68	16	241

Viewed (geographical distribution)

Total article views: 218 (including HTML, PDF, and XML) Thereof 218 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 30 Jan 2026

Short summary

We tested five tools that simulate rainfall to study how water affects soil. Because these tools differ in size and performance, we measured how evenly they spray and how strong the falling drops are. We found that each tool produces distinct rain patterns, but all can support soil studies when properly checked. Our work helps researchers choose practical and reliable equipment, especially when working in remote locations.


Total:	0
HTML:	0
PDF:	0
XML:	0