Influence of drop size distribution and kinetic energy in precipitation modelling for laboratory rainfall simulators

Ramli, Harris; Mohd Yusoff, Siti Aimi Nadia; Azmi, Mastura; Sabtu, Nuridah; Hezmi, Muhd Azril

doi:https://doi.org/10.5194/hess-2021-462

Preprints

https://doi.org/10.5194/hess-2021-462

Preprints

10 Nov 2021

| 10 Nov 2021

Status: this preprint was under review for the journal HESS but the revision was not accepted.

Influence of drop size distribution and kinetic energy in precipitation modelling for laboratory rainfall simulators

Harris Ramli, Siti Aimi Nadia Mohd Yusoff, Mastura Azmi, Nuridah Sabtu, and Muhd Azril Hezmi

Abstract. It is difficult to define the hydrologic and hydraulic characteristics of rain for research purposes, especially when trying to replicate natural rainfall using artificial rain on a small laboratory scale model. The aim of this paper was to use a drip-type rainfall simulator to design, build, calibrate, and run a simulated rainfall. Rainfall intensities of 40, 60 and 80 mm/h were used to represent heavy rainfall events of 1-hour duration. Flour pellet methods were used to obtain the drop size distribution of the simulated rainfall. The results show that the average drop size for all investigated rainfall intensities ranges from 3.0–3.4 mm. The median value of the drop size distribution or known as D₅₀ of simulated rainfall for 40, 60 and 80 mm/h are 3.4, 3.6, and 3.7 mm, respectively. Due to the comparatively low drop height (1.5 m), the terminal velocities monitored were between 63–75 % (8.45–8.65 m/s), which is lower than the value for natural rainfall with more than 90 % for terminal velocities. This condition also reduces rainfall kinetic energy of 25.88–28.51 J/m²mm compared to natural rainfall. This phenomenon is relatively common in portable rainfall simulators, representing the best exchange between all relevant rainfall parameters obtained with the given simulator set-up. Since the rainfall can be controlled, the erratic and unpredictable changeability of natural rainfall is eliminated. Emanating from the findings, drip-types rainfall simulator produces rainfall characteristics almost similar to natural rainfall-like characteristic is the main target.

Received: 10 Sep 2021 – Discussion started: 10 Nov 2021

Download & links

Harris Ramli et al.

Status: closed

RC1: 'Comment on hess-2021-462', Rolf Hut, 07 Dec 2021

Review of HESS-2021-462: ‘Influence of drop size distribution and kinetic energy in precipitation modelling for laboratory rainfall simulators’, by Ramli et. al.

Review by Rolf Hut.

The authors present the results of a series of experiments to test a newly build laboratory rainfall simulator. Rainfall simulators are an important tool in the bag of tools of (experimental) hydrologist and therefore a new rainfall simulator that has (potential) benefits over those already available is of interest and use to the readership of HESS. However, the article as presented does not allow, in its current form the readers of HESS to use the results / the rainfall simulator discussed in this work. Major revision and/or rewriting of the article is needed to clear up parts that are unclear and to add information that is currently missing.

Major issues

The authors claim that in their article ‘A rainfall simulator was designed, calibrated and tested’ (page 11, line 260, opening sentence of the Conclusion). However, the design as mentioned in paragraph 2 is not enough for a trained hydrologist to reproduce the system.
1. A photo of the system needs to be added.
2. The text mentions that every lateral contains 10 drip emitter at 0.8 meter, which would make the system 8 meters long. I’m assuming that this is 0.08m, as per the drawing in figure 1b.
3. The system seems to be 80 x 80 cm of nozzles. The text says that the flour pallet method is used to determine drop sizes. This is the main measurement presented in the paper. This method uses a 1m x 1x board (as the authors report). Rainfall simulators based on nozzles have notorious ‘edge effects’ where the rainfall characteristics are noticeably different at the edges of the simulator compared to the center. The authors make no mention of this at all. Open questions include:
  1. Did all the rain fall within the 1 square meter of the board? Of so: was it uniform? Their reporting of EU seems to suggest that it is, but it is unclear how EU was determined (see below).
  2. How was the calculation from water flow to mm/hour done? This is unclear from the text and has major impacts on the results and the conclusions.
4. Figure 1a suggest that the ‘soil’ is uneven and thus that different heights are possible in this system. If this is true it needs to be made explicit. If it is not, figure 1a needs to be corrected.
The literature contains many examples of rainfall simulators and the authors cite a fair few of these. These simulators cover a wide range of both surface areas and rainfall intensities targeted. This ranges from simulators of single drops, lab setups of the meter-scale and full multi-acre setups to be used on fields. The intensity ranges from drizzles to monsoons. The authors do not mention in their introduction what range their new simulator targets, nor for which applications it is build. Making this clear would help the readership with understanding the design choices.
Paragraph 3 contains a mix of theory and experimental setup that is hard to disentangle. The choices made in how to conduct the experiment are mentioned in between citations to literature. I strongly recommend separating paragraph 3 in a ‘theory’ and a ‘experimental setup’ paragraph. As I understand it (but I could be wrong) the following experiments are done:
1. To check the amount of water coming out of the system the flow meter is read for different settings of the pump pressure. It is not mentioned how this experiment was conducted: for how long was the valve opened? Was the flow allowed to settled before starting the measurement? Or was it the same 1 to 4 seconds mentioned in paragraph 3.2?
2. To check the drop sizes generated with this system, the flour pallet method was used with the flow opened for 1 to 4 seconds (how much? How was this determined? Was the idea to only have a few drops? How many?) Subsequently the flour drops were collected. From line 158 I conclude that 100 drops per intensity rate were collected. Table 1 shows that the system generates 113 drops per minute (at 40 mm/hour) so collecting would require just less than a minute. This does not rime with the 1 to 4 seconds. This needs clearing up on what was done and how it relates to what is reported. Related to this: table 1 says that the system generated 6.44 ml/minute at 40 mm/hour. 6.44 ml/minute is 386.4 ml/hour. On a 1 square meter plate, this would be 0.386 mm/hour. I can not combine all these numbers and make sense of them and the authors need to clarify this by clearly stating what experiments they did and how they processed the measurement results into their findings reported.
The authors use Foote & Du Toit (1969), their equation 3, to relate the drop sizes that they measure to terminal velocity. They use this terminal velocity to determine the maximum kinetic energy of the drops. They compare this to the actual kinetic energy but I do not understand how they calculate kinetic energy from their measured drop sizes. If they use one of the equations they mention, they are in a circular reasoning because all these relations are based on terminal velocity assumptions. If it is not, then the authors have to make clear how they calculate terminal velocity.
The drop size reported in the results and in figure 4 are concluded to be similar to natural rainfall (line 264). However, given my understanding of real rainfall, drop size distribution follow a Poisson distribution (by approximation) and the reported results do not. I suggest the authors look at the work of Remko Uijlenhoet on the relation between rain intensity and drop size distribution. The authors can add the expected drop size distribution for the given rain rate in their resulting figures. Based on the fit between these they can substantiate their claim that their rainfall simulator
One of the critical parameters for rainfall simulators is how uniform they rain over their area and in time. The authors report on EU, but as I understand it, they look at EU per nozzle and not distributed. For their simulator it is important to know if every nozzle has the same output (with regard to drop size distribution and flow rate). Given issues like pressure drop over the length of the laterals this might not be the case. Also the stability in time is an issue: the authors indicate they open and close the valve for 1 to 4 seconds. Is this how it will always be operated? Or will it be opened fully when used in practice? Because there might be start and stop issues in that case with the pressure fluctuating thus creating different sized drops. The authors would need to clarify this.

Given the issues above, I suggest a major re-write of the article with the additional information requested above added to it.

Rolf Hut

Citation: https://doi.org/10.5194/hess-2021-462-RC1

AC1: 'Reply on RC1', Mastura Azmi, 03 Feb 2022

The authors wish to thank the editorial team for managing the technical review. The authors also thank the reviewers for the comments and appreciate their support in making the paper more readable and more beneficial. Attached is the response to the comments and suggestions made by the reviewer (in PDF format).

No.	Comment	Action	Notes
1	The authors are kindly advised to revise the article as the article as presented does not allow, in its current form the readers of HESS to use the results / the rainfall simulator discussed in this work. Major revision and/or rewriting of the article is needed to clear up parts that are unclear and to add information that is currently missing.	They are considered in the revised manuscript.	The authors thank you for the comments and appreciate the support and involvement of the reviewers to improve the paper.
2	A photo of the system needs to be added. The text mentions that every lateral contains 10 drip emitter at 0.8 meter, which would make the system 8 meters long. I'm assuming that this is 0.08m, as per the drawing in figure 1b.	Regarding this comment, the author changed "0.8 m" to "8 cm", similar to the drawing in Figure 1b.	Thanks to the reviewer for this comment.
3	The system seems to be 80 x 80 cm of nozzles. The text says that the flour pallet method is used to determine drop sizes.	Regarding this comment, the author rewrites the sentence to provide a more clear view, where the sentences have been modified to read as follows: The system consists of 100 nozzle drip points distributed throughout a 72 cm × 72 cm square area. Ten lateral lines connected to the mainline contain ten nozzle drips equally spaced at 8 cm. The system's effective test area is 80 cm × 80 cm.	The sentences have been revised to avoid confusion with Figure 1b, as the lateral flow line spans over one metre (72cm + 39cm), with the drip spaced at an 8cm interval. Section 3.2 Drop Sizes elaborated on the method of using a flour pallet.
4	The author make no mention of the 'edge effects'? Did all the rain fall within the 1 square meter of the board? Of so: was it uniform?	Regarding this comment, the author added a paragraph on rainfall simulator calibration to provide a more clear view: The rainfall simulator was calibrated in terms of rainfall intensity to achieve reproducible and consistent rainfall characteristics. Throughout the calibration and experiment, any element that may affect the changes in wind flow, such as an air conditioner or fan, is controlled to ensure that the raindrop falls vertically within the effective test area. The calibrating process was split into two parts. The first step measured the rainfall intensity and spatial rainfall distribution on the designated surface. To obtain high-resolution datasets, 100 cylindrical rainfall collectors are positioned on an area of 80 cm x 80 cm under the drip used. Each collector was weighed to determine the amount of rain collected after one hour of simulated rainfall. Based on the observations, no droplet crosses the cylindrical during the calibration and all land directly on the cylindrical beneath the drip. This initial calibration step ensures that each nozzle produces an equal amount of rainfall. A second step involved using a single large, plot-sized collection to determine net rainfall intensities. The volumetric method of flow measurement was utilised to calibrate the simulated real intensity. Two laboratory steel trays with top dimensions of 80cm x 80cm x 10cm and bottom dimensions of 100cm x 100cm x 8cm were used for volume control and placed beneath drip systems. The different heights of the top and bottom are merely a coincidence due to the laboratory's available tray. The primary technical requirement is that the central collection tray (small) must fit within the dimensions of the designed test area and be tall enough to collect the intended rainfall intensity (in this work, the maximum is 80 mm/h.), whereas the secondary collection tray must be larger than the effective test area. Collector boxes were placed in a central location (in relation to the drip location) and collected the precipitated volume at a set pressure. The volume of precipitated water was determined using a measuring cylinder. A ruler was used to measure the water level, and then the precipitation volume collected was recorded. No raindrops landed on the large bottom tray during the second calibration stage based on the observations. This demonstrates that the raindrop area's uniformity is reproducible.	Thanks to the reviewer for this comment and suggestion.
5	How was the calculation from water flow to mm/hour done?	Regarding this comment, the author added sentences to provide a more clear view: The calibration result enables the flowmeter to correlate the amount of water controlled by the flowmeter to the amount of water emitted from the nozzle. Thus, the simulator delivers the desired rainfall intensity (mm/hr).	Thanks to the reviewer for this comment and suggestion.
6	Figure 1a suggest that the 'soil' is uneven and thus that different heights are possible in this system. If this is true it needs to be made explicit. If it is not, figure 1a needs to be corrected	No changes in manuscript	Thanks to the reviewer for this comment. The uneven soil profiles result from another aspect of this research that is not covered in this paper. Calibration results indicated that small changes to these soil profiles do not affect the rainfall simulator's performance.
7	The authors do not mention in their introduction what range their new simulator targets, nor for which applications it is build.	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. The main intention of this simulator setup is to simulate targeted rainfall intensities similar to the cases of landslide occurrences in the authors country. The simulator is applied to a laboratory setup of a specific selection of slope criteria for further understanding on rainfall impact in groundwater level changes. It is beyond this manuscript context; therefore, the authors excluded the explanation of the simulator's application.
8	Paragraph 3 contains a mix of theory and experimental setup that is hard to disentangle. The choices made in how to conduct the experiment are mentioned in between citations to literature. I strongly recommend separating paragraph 3 in a 'theory' and a 'experimental setup' paragraph.	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. The third paragraph discussed solely with the subject's literature.
9	To check the amount of water coming out of the system the flow meter is read for different settings of the pump pressure. It is not mentioned how this experiment was conducted: for how long was the valve opened? Was the flow allowed to settled before starting the measurement? Or was it the same 1 to 4 seconds mentioned in paragraph 3.2?	As mentioned in the previous comment (No. 4 and No. 5), the author added a paragraph on rainfall simulator calibration to provide a clearer view.	Thanks to the reviewer for this comment and suggestion.
10	To check the drop sizes generated with this system, the flour pallet method was used with the flow opened for 1 to 4 seconds (how much? How was this determined? Was the idea to only have a few drops? How many?) This needs clearing up on what was done and how it relates to what is reported.	Regarding this comment, the author added sentences to provide a more clear view: According to Kathiravelu et al., 2016, the flour pallet test would be conducted by various researchers between 1 and 4 seconds, depending on the intensity of rainfall. In our experiment, a 1.0 m x 1.0 m plate containing a 2.54 cm (1 inch) layer of uncompacted fine wheat flour was exposed to a split second of rainfall. The drops must not fall at the same point during rainfall simulation and formed corresponding to the drop size on impact. The objective is to collect a single drop of water. The flour plate was positioned 1.5 metres below the drip and covered with two layers of tray. A collection tray is used on both the top and bottom trays. The top collection tray's purpose is to capture the initial 1 to 2 seconds of rainfall when the valve is opened. This ensures that every nozzle produces a raindrop. Simultaneously with the top tray being set aside, the flour plate was exposed for a split second before being covered by the bottom tray. The valve was immediately closed to halt the rain. Based on the observation, the raindrop pattern is within the effective test area.	Thanks to the reviewer for this comment and suggestion.
11	The authors indicate they open and close the valve for 1 to 4 seconds. Is this how it will always be operated? Or will it be opened fully when used in practice?	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. As explained in comment No. 10, the statement of 1 to 4 seconds is for the flour pallet test only. During the actual rainfall simulation test, the intended volume of rainfall is applied according to the rainfall intensity and duration.

Citation: https://doi.org/10.5194/hess-2021-462-AC1

RC2: 'Comment on hess-2021-462', Luca G. Lanza, 02 Jan 2022

The paper addresses the performance of a designed laboratory rainfall simulator for use in soil erosion and other experiments. Reproducing rainfall characteristics with accuracy in the laboratory is a requisite to support some experimental hydrological studies under rigorously controlled conditions. This would also avoid dealing with the inherent variability of the rainfall process and the many sources of uncertainty affecting any real-word experiment in the field.

In this regard, the topic of this manuscript is relevant and surely of interest within the hydrological sciences, and therefore suited to the journal.

However, the authors first fail to show any improvement of their proposed solution over those already available in the literature, starting from the ones presented in papers that are mentioned among the references of this manuscript. There is no evidence in the manuscript, indeed, that the presented setup performs any better than very similar solutions that already exists.

Second, the proposed setup for the rain simulator is presented in a rather superficial manner, with scarce quantitative indication of the actual performance of the device. It seems that only the number and diameter of the generated drops are measured for validation, while all other parameters (including the fall velocity) are derived from empirical formulations reported in the literature for real rainfall events.

Most important, little effort is presented to demonstrate that the few measured quantities (basically the drop size distribution – DSD – of the generated rainfall) is like the one expected from real rainfall events (at least in the declared range of rainfall intensity generated by the device). Comparison with typical DSD associated with the generated rain intensity values under various meteo-climatic conditions is totally missing.

Finally, the main comments reported in the conclusions do not reflect results of the present manuscript (e.g., no calibration exercise is described in the text, no validation with real rainfall characteristics is provided, etc.). General statements about the dependence of the drop size distribution and the kinetic energy of the raindrops on the rainfall intensity are declared as results of the present study, while obviously they are not.

My knowledge of the English language is rather poor, and I’m not qualified to make a revision of the English used, but my feeling is that the wording and grammar throughout the text should be extensively checked and corrected. Also, the text is quite repetitive in many sections and the presentation should be largely improved

Based on the above considerations, I believe that the manuscript is not ready for publication on HESS, and needs to be largely revised, not only in the presentation of the work but also in its content (first checking the real improvement provided over previously published research) before it can be accepted.

Minor comments are provided in the attached pdf.

Citation: https://doi.org/10.5194/hess-2021-462-RC2

AC2: 'Reply on RC2', Mastura Azmi, 03 Feb 2022

No.	Comment	Action	Notes
1	The authors are kindly advised to revise the article as the authors first fail to show any improvement of their proposed solution over those already available in the literature, starting from the ones presented in papers that are mentioned among the references of this manuscript. There is no evidence in the manuscript, indeed, that the presented setup performs any better than very similar solutions that already exists.	They are considered in the revised manuscript.	The authors thank you for the comments and appreciate the support and involvement of the reviewers to improve the paper. The novelty of this paper compared to Mendes (2021), is that this paper proves that the usage of drip emitter for a high rainfall intensity is more suitable than spray nozzle. This paper also proves that raindrops' size distribution is more consistent when using drip emitter. This can be seen from consistent median value in the normal distribution of D₅₀. This paper also proves that the collection area is important in rainfall simulator selection (drip emitter vs. spray nozzle).
2	The main comments reported in the conclusions do not reflect results of the present manuscript	Terminal velocity is calculated based on the equation (3) – Text in abstract needs to change from monitored to calculated RC2: the main comments reported in the conclusions do not reflect results of the present manuscript (e.g., no calibration exercise is described in the text, no validation with real rainfall characteristics is provided, etc.). General statements about the dependence of the drop size distribution and the kinetic energy of the raindrops on the rainfall intensity are declared as results of the present study, while obviously they are not.
3	This is correctly states as a requirement but not demonstrated in the manuscript. "Surely, simulated rainfall should made to be as close to natural rainfall as possible, especially on droplets' fall velocity and size distribution."	No changes in manuscript	Thanks to the reviewer for this comment. The main novelty of our paper is that we prioritise improving the simulated rainfall droplet size distribution, as this is something that other researchers can effectively replicate in comparison to the fall velocity parameter.
4	Correction of the terminology "…intensity of the raindrop…"	Regarding this comment, the author had changed "raindrop" to "rainfall".	Thanks to the reviewer for this comment and suggestion.
5	This is not demonstrate in the manuscript. "…capable of producing droplet characteristics identical to natural rainfall."	Regarding to this comment, the author adds more detailed information on this sentence to give a clearer view, where the sentences has been modified to: "…capable of producing droplet characteristics identical to those found in natural rainfall from a drop size distribution perspective."	Thanks to the reviewer for this comment and suggestion.
6	Correction of the terminology. "The pipe diameter is 2 cm while the spray diameter is 4 mm."	Regarding this comment, the author changed "spray" to "nozzle".	Thanks to the reviewer for this comment and suggestion.
7	It is also important to note that high rainfall intensities degrade the quality of measurements using the disdrometers or imaging techniques, due to background noise (Mendes et al., 2021).	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. This sentence intends to emphasise that this study, like Mendes et al. 2021, is not suitable for using disdrometers due to the high rainfall intensity (40 mm/hr to 80 mm/hr).
8	Was this calibrated in some way or is it a literature result? D_d = 14.56M_f^0.354and D_d (mm) = 0.985D_f (mm)^1.02	Regarding this comment, the author omitted the sentences: The relationship between the diameter of the drop, D_d (in mm), and the mass of the flour ball, M_f (in mg), is D_d = 14.56M_f^0.354, respectively. To estimate the drop diameter as a function of the flour ball diameter D_d (mm) = 0.985D_f (mm)^1.02, the diameter of flour balls can be considered practically equivalent to that of the drops.	Thanks to the reviewer for this comment and suggestion. The sentences are omitted because the formula is a literature result based on Navas et al., 1990. In our case, we might not need to use the formula since the drop diameter is measured, and the result shows a consistent median value in the normal distribution of D₅₀.
9	No measurements of fall velocity and kinetic energy are reported	Regarding this comment, the author change the typo “I = Rainfall Intensity,” to the “I = Rainfall Intensity.” and missing words added to the sentences: I = Rainfall Intensity. Kinetic energy can also be measured from the physical properties of simulated rainfall: raindrop size, fall velocity and drop size distribution.	Thanks to the reviewer for this comment and suggestion.
10	Why is this different from Eq. 3 used above ?	Regarding this comment, the Equation is same as the Eq.3	Thanks to the reviewer for this comment and suggestion.
11	Could not see any calibration exercise described in this paper. What is the meaning of "calibrated" here ?	Regarding this comment, the author added a paragraph on rainfall simulator calibration to provide a more clear view: The rainfall simulator was calibrated in terms of rainfall intensity to achieve reproducible and consistent rainfall characteristics. Throughout the calibration and experiment, any element that may affect the changes in wind flow, such as an air conditioner or fan, is controlled to ensure that the raindrop falls vertically within the effective test area. The calibrating process was split into two parts. The first step measured the rainfall intensity and spatial rainfall distribution on the designated surface. To obtain high-resolution datasets, 100 cylindrical rainfall collectors are positioned on an area of 80 cm x 80 cm under the drip used. Each collector was weighed to determine the amount of rain collected after one hour of simulated rainfall. Based on the observations, no droplet crosses the cylindrical during the calibration and all land directly on the cylindrical beneath the drip. This initial calibration step ensures that each nozzle produces an equal amount of rainfall. A second step involved using a single large, plot-sized collection to determine net rainfall intensities. The volumetric method of flow measurement was utilised to calibrate the simulated real intensity. Two laboratory steel trays with top dimensions of 80cm x 80cm x 10cm and bottom dimensions of 100cm x 100cm x 8cm were used for volume control and placed beneath drip systems. The different heights of the top and bottom are merely a coincidence due to the laboratory's available tray. The primary technical requirement is that the central collection tray (small) must fit within the dimensions of the designed test area and be tall enough to collect the intended rainfall intensity (in this work, the maximum is 80 mm/h.), whereas the secondary collection tray must be larger than the effective test area. Collector boxes were placed in a central location (in relation to the drip location) and collected the precipitated volume at a set pressure. The volume of precipitated water was determined using a measuring cylinder. A ruler was used to measure the water level, and then the precipitation volume collected was recorded. No raindrops landed on the large bottom tray during the second calibration stage based on the observations. This demonstrates that the raindrop area's uniformity is reproducible.	Thanks to the reviewer for this comment and suggestion.
12	Repeated sentence	Regarding this comment, the author omitted a sentence to avoid repetition in conclusion: The drop size distributions were obtained by using a flour pellet methodology, described by Kathiravelu et.al (2016).	Thanks to the reviewer for this comment and suggestion.
13	This was not demonstrated in the paper.	Regarding this comment, the author added a sentence to provide a more specific conclusion: …natural rainfall is achieved from the perspective of rainfall intensity and drop size distribution.	Thanks to the reviewer for this comment and suggestion.
14	The results presented in the manuscript do not show that. It was simply assumed from literature results.	Regarding this comment, the author rewrite a sentence to provide a more specific conclusion: This study discovered that DSD might affect both kinetic energy and terminal velocity.	The sentences have been rewritten so that they are less ambiguous.

Citation: https://doi.org/10.5194/hess-2021-462-AC2

Status: closed

RC1: 'Comment on hess-2021-462', Rolf Hut, 07 Dec 2021

Review of HESS-2021-462: ‘Influence of drop size distribution and kinetic energy in precipitation modelling for laboratory rainfall simulators’, by Ramli et. al.

Review by Rolf Hut.

Major issues

The authors claim that in their article ‘A rainfall simulator was designed, calibrated and tested’ (page 11, line 260, opening sentence of the Conclusion). However, the design as mentioned in paragraph 2 is not enough for a trained hydrologist to reproduce the system.
1. A photo of the system needs to be added.
2. The text mentions that every lateral contains 10 drip emitter at 0.8 meter, which would make the system 8 meters long. I’m assuming that this is 0.08m, as per the drawing in figure 1b.
3. The system seems to be 80 x 80 cm of nozzles. The text says that the flour pallet method is used to determine drop sizes. This is the main measurement presented in the paper. This method uses a 1m x 1x board (as the authors report). Rainfall simulators based on nozzles have notorious ‘edge effects’ where the rainfall characteristics are noticeably different at the edges of the simulator compared to the center. The authors make no mention of this at all. Open questions include:
  1. Did all the rain fall within the 1 square meter of the board? Of so: was it uniform? Their reporting of EU seems to suggest that it is, but it is unclear how EU was determined (see below).
  2. How was the calculation from water flow to mm/hour done? This is unclear from the text and has major impacts on the results and the conclusions.
4. Figure 1a suggest that the ‘soil’ is uneven and thus that different heights are possible in this system. If this is true it needs to be made explicit. If it is not, figure 1a needs to be corrected.
The literature contains many examples of rainfall simulators and the authors cite a fair few of these. These simulators cover a wide range of both surface areas and rainfall intensities targeted. This ranges from simulators of single drops, lab setups of the meter-scale and full multi-acre setups to be used on fields. The intensity ranges from drizzles to monsoons. The authors do not mention in their introduction what range their new simulator targets, nor for which applications it is build. Making this clear would help the readership with understanding the design choices.
Paragraph 3 contains a mix of theory and experimental setup that is hard to disentangle. The choices made in how to conduct the experiment are mentioned in between citations to literature. I strongly recommend separating paragraph 3 in a ‘theory’ and a ‘experimental setup’ paragraph. As I understand it (but I could be wrong) the following experiments are done:
1. To check the amount of water coming out of the system the flow meter is read for different settings of the pump pressure. It is not mentioned how this experiment was conducted: for how long was the valve opened? Was the flow allowed to settled before starting the measurement? Or was it the same 1 to 4 seconds mentioned in paragraph 3.2?
2. To check the drop sizes generated with this system, the flour pallet method was used with the flow opened for 1 to 4 seconds (how much? How was this determined? Was the idea to only have a few drops? How many?) Subsequently the flour drops were collected. From line 158 I conclude that 100 drops per intensity rate were collected. Table 1 shows that the system generates 113 drops per minute (at 40 mm/hour) so collecting would require just less than a minute. This does not rime with the 1 to 4 seconds. This needs clearing up on what was done and how it relates to what is reported. Related to this: table 1 says that the system generated 6.44 ml/minute at 40 mm/hour. 6.44 ml/minute is 386.4 ml/hour. On a 1 square meter plate, this would be 0.386 mm/hour. I can not combine all these numbers and make sense of them and the authors need to clarify this by clearly stating what experiments they did and how they processed the measurement results into their findings reported.
The authors use Foote & Du Toit (1969), their equation 3, to relate the drop sizes that they measure to terminal velocity. They use this terminal velocity to determine the maximum kinetic energy of the drops. They compare this to the actual kinetic energy but I do not understand how they calculate kinetic energy from their measured drop sizes. If they use one of the equations they mention, they are in a circular reasoning because all these relations are based on terminal velocity assumptions. If it is not, then the authors have to make clear how they calculate terminal velocity.
The drop size reported in the results and in figure 4 are concluded to be similar to natural rainfall (line 264). However, given my understanding of real rainfall, drop size distribution follow a Poisson distribution (by approximation) and the reported results do not. I suggest the authors look at the work of Remko Uijlenhoet on the relation between rain intensity and drop size distribution. The authors can add the expected drop size distribution for the given rain rate in their resulting figures. Based on the fit between these they can substantiate their claim that their rainfall simulator
One of the critical parameters for rainfall simulators is how uniform they rain over their area and in time. The authors report on EU, but as I understand it, they look at EU per nozzle and not distributed. For their simulator it is important to know if every nozzle has the same output (with regard to drop size distribution and flow rate). Given issues like pressure drop over the length of the laterals this might not be the case. Also the stability in time is an issue: the authors indicate they open and close the valve for 1 to 4 seconds. Is this how it will always be operated? Or will it be opened fully when used in practice? Because there might be start and stop issues in that case with the pressure fluctuating thus creating different sized drops. The authors would need to clarify this.

Given the issues above, I suggest a major re-write of the article with the additional information requested above added to it.

Rolf Hut

Citation: https://doi.org/10.5194/hess-2021-462-RC1

AC1: 'Reply on RC1', Mastura Azmi, 03 Feb 2022

No.	Comment	Action	Notes
1	The authors are kindly advised to revise the article as the article as presented does not allow, in its current form the readers of HESS to use the results / the rainfall simulator discussed in this work. Major revision and/or rewriting of the article is needed to clear up parts that are unclear and to add information that is currently missing.	They are considered in the revised manuscript.	The authors thank you for the comments and appreciate the support and involvement of the reviewers to improve the paper.
2	A photo of the system needs to be added. The text mentions that every lateral contains 10 drip emitter at 0.8 meter, which would make the system 8 meters long. I'm assuming that this is 0.08m, as per the drawing in figure 1b.	Regarding this comment, the author changed "0.8 m" to "8 cm", similar to the drawing in Figure 1b.	Thanks to the reviewer for this comment.
3	The system seems to be 80 x 80 cm of nozzles. The text says that the flour pallet method is used to determine drop sizes.	Regarding this comment, the author rewrites the sentence to provide a more clear view, where the sentences have been modified to read as follows: The system consists of 100 nozzle drip points distributed throughout a 72 cm × 72 cm square area. Ten lateral lines connected to the mainline contain ten nozzle drips equally spaced at 8 cm. The system's effective test area is 80 cm × 80 cm.	The sentences have been revised to avoid confusion with Figure 1b, as the lateral flow line spans over one metre (72cm + 39cm), with the drip spaced at an 8cm interval. Section 3.2 Drop Sizes elaborated on the method of using a flour pallet.
4	The author make no mention of the 'edge effects'? Did all the rain fall within the 1 square meter of the board? Of so: was it uniform?	Regarding this comment, the author added a paragraph on rainfall simulator calibration to provide a more clear view: The rainfall simulator was calibrated in terms of rainfall intensity to achieve reproducible and consistent rainfall characteristics. Throughout the calibration and experiment, any element that may affect the changes in wind flow, such as an air conditioner or fan, is controlled to ensure that the raindrop falls vertically within the effective test area. The calibrating process was split into two parts. The first step measured the rainfall intensity and spatial rainfall distribution on the designated surface. To obtain high-resolution datasets, 100 cylindrical rainfall collectors are positioned on an area of 80 cm x 80 cm under the drip used. Each collector was weighed to determine the amount of rain collected after one hour of simulated rainfall. Based on the observations, no droplet crosses the cylindrical during the calibration and all land directly on the cylindrical beneath the drip. This initial calibration step ensures that each nozzle produces an equal amount of rainfall. A second step involved using a single large, plot-sized collection to determine net rainfall intensities. The volumetric method of flow measurement was utilised to calibrate the simulated real intensity. Two laboratory steel trays with top dimensions of 80cm x 80cm x 10cm and bottom dimensions of 100cm x 100cm x 8cm were used for volume control and placed beneath drip systems. The different heights of the top and bottom are merely a coincidence due to the laboratory's available tray. The primary technical requirement is that the central collection tray (small) must fit within the dimensions of the designed test area and be tall enough to collect the intended rainfall intensity (in this work, the maximum is 80 mm/h.), whereas the secondary collection tray must be larger than the effective test area. Collector boxes were placed in a central location (in relation to the drip location) and collected the precipitated volume at a set pressure. The volume of precipitated water was determined using a measuring cylinder. A ruler was used to measure the water level, and then the precipitation volume collected was recorded. No raindrops landed on the large bottom tray during the second calibration stage based on the observations. This demonstrates that the raindrop area's uniformity is reproducible.	Thanks to the reviewer for this comment and suggestion.
5	How was the calculation from water flow to mm/hour done?	Regarding this comment, the author added sentences to provide a more clear view: The calibration result enables the flowmeter to correlate the amount of water controlled by the flowmeter to the amount of water emitted from the nozzle. Thus, the simulator delivers the desired rainfall intensity (mm/hr).	Thanks to the reviewer for this comment and suggestion.
6	Figure 1a suggest that the 'soil' is uneven and thus that different heights are possible in this system. If this is true it needs to be made explicit. If it is not, figure 1a needs to be corrected	No changes in manuscript	Thanks to the reviewer for this comment. The uneven soil profiles result from another aspect of this research that is not covered in this paper. Calibration results indicated that small changes to these soil profiles do not affect the rainfall simulator's performance.
7	The authors do not mention in their introduction what range their new simulator targets, nor for which applications it is build.	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. The main intention of this simulator setup is to simulate targeted rainfall intensities similar to the cases of landslide occurrences in the authors country. The simulator is applied to a laboratory setup of a specific selection of slope criteria for further understanding on rainfall impact in groundwater level changes. It is beyond this manuscript context; therefore, the authors excluded the explanation of the simulator's application.
8	Paragraph 3 contains a mix of theory and experimental setup that is hard to disentangle. The choices made in how to conduct the experiment are mentioned in between citations to literature. I strongly recommend separating paragraph 3 in a 'theory' and a 'experimental setup' paragraph.	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. The third paragraph discussed solely with the subject's literature.
9	To check the amount of water coming out of the system the flow meter is read for different settings of the pump pressure. It is not mentioned how this experiment was conducted: for how long was the valve opened? Was the flow allowed to settled before starting the measurement? Or was it the same 1 to 4 seconds mentioned in paragraph 3.2?	As mentioned in the previous comment (No. 4 and No. 5), the author added a paragraph on rainfall simulator calibration to provide a clearer view.	Thanks to the reviewer for this comment and suggestion.
10	To check the drop sizes generated with this system, the flour pallet method was used with the flow opened for 1 to 4 seconds (how much? How was this determined? Was the idea to only have a few drops? How many?) This needs clearing up on what was done and how it relates to what is reported.	Regarding this comment, the author added sentences to provide a more clear view: According to Kathiravelu et al., 2016, the flour pallet test would be conducted by various researchers between 1 and 4 seconds, depending on the intensity of rainfall. In our experiment, a 1.0 m x 1.0 m plate containing a 2.54 cm (1 inch) layer of uncompacted fine wheat flour was exposed to a split second of rainfall. The drops must not fall at the same point during rainfall simulation and formed corresponding to the drop size on impact. The objective is to collect a single drop of water. The flour plate was positioned 1.5 metres below the drip and covered with two layers of tray. A collection tray is used on both the top and bottom trays. The top collection tray's purpose is to capture the initial 1 to 2 seconds of rainfall when the valve is opened. This ensures that every nozzle produces a raindrop. Simultaneously with the top tray being set aside, the flour plate was exposed for a split second before being covered by the bottom tray. The valve was immediately closed to halt the rain. Based on the observation, the raindrop pattern is within the effective test area.	Thanks to the reviewer for this comment and suggestion.
11	The authors indicate they open and close the valve for 1 to 4 seconds. Is this how it will always be operated? Or will it be opened fully when used in practice?	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. As explained in comment No. 10, the statement of 1 to 4 seconds is for the flour pallet test only. During the actual rainfall simulation test, the intended volume of rainfall is applied according to the rainfall intensity and duration.

Citation: https://doi.org/10.5194/hess-2021-462-AC1

RC2: 'Comment on hess-2021-462', Luca G. Lanza, 02 Jan 2022

In this regard, the topic of this manuscript is relevant and surely of interest within the hydrological sciences, and therefore suited to the journal.

Minor comments are provided in the attached pdf.

Citation: https://doi.org/10.5194/hess-2021-462-RC2

AC2: 'Reply on RC2', Mastura Azmi, 03 Feb 2022

No.	Comment	Action	Notes
1	The authors are kindly advised to revise the article as the authors first fail to show any improvement of their proposed solution over those already available in the literature, starting from the ones presented in papers that are mentioned among the references of this manuscript. There is no evidence in the manuscript, indeed, that the presented setup performs any better than very similar solutions that already exists.	They are considered in the revised manuscript.	The authors thank you for the comments and appreciate the support and involvement of the reviewers to improve the paper. The novelty of this paper compared to Mendes (2021), is that this paper proves that the usage of drip emitter for a high rainfall intensity is more suitable than spray nozzle. This paper also proves that raindrops' size distribution is more consistent when using drip emitter. This can be seen from consistent median value in the normal distribution of D₅₀. This paper also proves that the collection area is important in rainfall simulator selection (drip emitter vs. spray nozzle).
2	The main comments reported in the conclusions do not reflect results of the present manuscript	Terminal velocity is calculated based on the equation (3) – Text in abstract needs to change from monitored to calculated RC2: the main comments reported in the conclusions do not reflect results of the present manuscript (e.g., no calibration exercise is described in the text, no validation with real rainfall characteristics is provided, etc.). General statements about the dependence of the drop size distribution and the kinetic energy of the raindrops on the rainfall intensity are declared as results of the present study, while obviously they are not.
3	This is correctly states as a requirement but not demonstrated in the manuscript. "Surely, simulated rainfall should made to be as close to natural rainfall as possible, especially on droplets' fall velocity and size distribution."	No changes in manuscript	Thanks to the reviewer for this comment. The main novelty of our paper is that we prioritise improving the simulated rainfall droplet size distribution, as this is something that other researchers can effectively replicate in comparison to the fall velocity parameter.
4	Correction of the terminology "…intensity of the raindrop…"	Regarding this comment, the author had changed "raindrop" to "rainfall".	Thanks to the reviewer for this comment and suggestion.
5	This is not demonstrate in the manuscript. "…capable of producing droplet characteristics identical to natural rainfall."	Regarding to this comment, the author adds more detailed information on this sentence to give a clearer view, where the sentences has been modified to: "…capable of producing droplet characteristics identical to those found in natural rainfall from a drop size distribution perspective."	Thanks to the reviewer for this comment and suggestion.
6	Correction of the terminology. "The pipe diameter is 2 cm while the spray diameter is 4 mm."	Regarding this comment, the author changed "spray" to "nozzle".	Thanks to the reviewer for this comment and suggestion.
7	It is also important to note that high rainfall intensities degrade the quality of measurements using the disdrometers or imaging techniques, due to background noise (Mendes et al., 2021).	No changes in manuscript	Thanks to the reviewer for this comment and suggestion. This sentence intends to emphasise that this study, like Mendes et al. 2021, is not suitable for using disdrometers due to the high rainfall intensity (40 mm/hr to 80 mm/hr).
8	Was this calibrated in some way or is it a literature result? D_d = 14.56M_f^0.354and D_d (mm) = 0.985D_f (mm)^1.02	Regarding this comment, the author omitted the sentences: The relationship between the diameter of the drop, D_d (in mm), and the mass of the flour ball, M_f (in mg), is D_d = 14.56M_f^0.354, respectively. To estimate the drop diameter as a function of the flour ball diameter D_d (mm) = 0.985D_f (mm)^1.02, the diameter of flour balls can be considered practically equivalent to that of the drops.	Thanks to the reviewer for this comment and suggestion. The sentences are omitted because the formula is a literature result based on Navas et al., 1990. In our case, we might not need to use the formula since the drop diameter is measured, and the result shows a consistent median value in the normal distribution of D₅₀.
9	No measurements of fall velocity and kinetic energy are reported	Regarding this comment, the author change the typo “I = Rainfall Intensity,” to the “I = Rainfall Intensity.” and missing words added to the sentences: I = Rainfall Intensity. Kinetic energy can also be measured from the physical properties of simulated rainfall: raindrop size, fall velocity and drop size distribution.	Thanks to the reviewer for this comment and suggestion.
10	Why is this different from Eq. 3 used above ?	Regarding this comment, the Equation is same as the Eq.3	Thanks to the reviewer for this comment and suggestion.
11	Could not see any calibration exercise described in this paper. What is the meaning of "calibrated" here ?	Regarding this comment, the author added a paragraph on rainfall simulator calibration to provide a more clear view: The rainfall simulator was calibrated in terms of rainfall intensity to achieve reproducible and consistent rainfall characteristics. Throughout the calibration and experiment, any element that may affect the changes in wind flow, such as an air conditioner or fan, is controlled to ensure that the raindrop falls vertically within the effective test area. The calibrating process was split into two parts. The first step measured the rainfall intensity and spatial rainfall distribution on the designated surface. To obtain high-resolution datasets, 100 cylindrical rainfall collectors are positioned on an area of 80 cm x 80 cm under the drip used. Each collector was weighed to determine the amount of rain collected after one hour of simulated rainfall. Based on the observations, no droplet crosses the cylindrical during the calibration and all land directly on the cylindrical beneath the drip. This initial calibration step ensures that each nozzle produces an equal amount of rainfall. A second step involved using a single large, plot-sized collection to determine net rainfall intensities. The volumetric method of flow measurement was utilised to calibrate the simulated real intensity. Two laboratory steel trays with top dimensions of 80cm x 80cm x 10cm and bottom dimensions of 100cm x 100cm x 8cm were used for volume control and placed beneath drip systems. The different heights of the top and bottom are merely a coincidence due to the laboratory's available tray. The primary technical requirement is that the central collection tray (small) must fit within the dimensions of the designed test area and be tall enough to collect the intended rainfall intensity (in this work, the maximum is 80 mm/h.), whereas the secondary collection tray must be larger than the effective test area. Collector boxes were placed in a central location (in relation to the drip location) and collected the precipitated volume at a set pressure. The volume of precipitated water was determined using a measuring cylinder. A ruler was used to measure the water level, and then the precipitation volume collected was recorded. No raindrops landed on the large bottom tray during the second calibration stage based on the observations. This demonstrates that the raindrop area's uniformity is reproducible.	Thanks to the reviewer for this comment and suggestion.
12	Repeated sentence	Regarding this comment, the author omitted a sentence to avoid repetition in conclusion: The drop size distributions were obtained by using a flour pellet methodology, described by Kathiravelu et.al (2016).	Thanks to the reviewer for this comment and suggestion.
13	This was not demonstrated in the paper.	Regarding this comment, the author added a sentence to provide a more specific conclusion: …natural rainfall is achieved from the perspective of rainfall intensity and drop size distribution.	Thanks to the reviewer for this comment and suggestion.
14	The results presented in the manuscript do not show that. It was simply assumed from literature results.	Regarding this comment, the author rewrite a sentence to provide a more specific conclusion: This study discovered that DSD might affect both kinetic energy and terminal velocity.	The sentences have been rewritten so that they are less ambiguous.

Citation: https://doi.org/10.5194/hess-2021-462-AC2

Harris Ramli et al.

Viewed

Total article views: 897 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
619	250	28	897	19	15

HTML: 619
PDF: 250
XML: 28
Total: 897
BibTeX: 19
EndNote: 15

Views and downloads (calculated since 10 Nov 2021)

Month	HTML	PDF	XML	Total
Nov 2021	216	34	5	255
Dec 2021	51	15	1	67
Jan 2022	49	18	3	70
Feb 2022	42	13	2	57
Mar 2022	25	11	3	39
Apr 2022	15	10	0	25
May 2022	12	9	2	23
Jun 2022	6	4	1	11
Jul 2022	25	2	0	27
Aug 2022	6	7	0	13
Sep 2022	14	11	0	25
Oct 2022	13	7	2	22
Nov 2022	12	4	0	16
Dec 2022	10	8	0	18
Jan 2023	9	30	0	39
Feb 2023	11	6	0	17
Mar 2023	9	5	1	15
Apr 2023	6	4	1	11
May 2023	9	5	0	14
Jun 2023	8	8	1	17
Jul 2023	38	12	1	51
Aug 2023	24	14	2	40
Sep 2023	9	13	3	25

Cumulative views and downloads (calculated since 10 Nov 2021)

Month	HTML	PDF	XML	Total
Nov 2021	216	34	5	255
Dec 2021	51	15	1	67
Jan 2022	49	18	3	70
Feb 2022	42	13	2	57
Mar 2022	25	11	3	39
Apr 2022	15	10	0	25
May 2022	12	9	2	23
Jun 2022	6	4	1	11
Jul 2022	25	2	0	27
Aug 2022	6	7	0	13
Sep 2022	14	11	0	25
Oct 2022	13	7	2	22
Nov 2022	12	4	0	16
Dec 2022	10	8	0	18
Jan 2023	9	30	0	39
Feb 2023	11	6	0	17
Mar 2023	9	5	1	15
Apr 2023	6	4	1	11
May 2023	9	5	0	14
Jun 2023	8	8	1	17
Jul 2023	38	12	1	51
Aug 2023	24	14	2	40
Sep 2023	9	13	3	25

Viewed (geographical distribution)

Total article views: 857 (including HTML, PDF, and XML) Thereof 857 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Sep 2023

Short summary

A rainfall simulator was designed, calibrated, and tested to produce heavy rainfall intensities like natural rainfall. Properties considered include appropriate density under constant head conditions that will influence the intensity of the raindrops based on the spacing of drop formers, uniformity of the application and the droplet size distribution. The results from this study show that kinetic energy and terminal velocity was dependent on particle size distribution.


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