the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Spatial accessibility of emergency medical services under inclement weather: A case study in Beijing, China
Yuting Zhang
Kai Liu
Xiaoyong Ni
Ming Wang
Jianchun Zheng
Mengting Liu
Dapeng Yu
Abstract. The accessibility of emergency medical services (EMSs) is not only determined by the distribution of emergency medical facilities but is also very vulnerable to weather conditions. Inclement weather could affect the efficiency of the city's traffic network and further affect the response time of EMSs, which could therefore be an essential impact factor on the safety of human lives. This study proposes an EMS-accessibility quantification method based on selected indicators and explores the influence of inclement weather on EMS accessibility and identifies the hot spots that have difficulty accessing timely EMSs. A case study was implemented in Beijing, which is a typical megacity in China, based on the ground-truth traffic data of the whole city in 2019. The results show that inclement weather has a general negative impact on EMS accessibility. The 15-min EMS coverage rate of the area could have a maximum reduction of 13 % at the citywide scale and could reach over 40 % in some suburban townships. Although on the whole, the urban area would have more traffic speed reduction, towns with lower baseline EMS accessibility is more vulnerable to inclement weather, furthermore, the proportion of elderly population in these towns is also higher than the average level of the whole city. Under the worst scenario in 2019, 12.6 % of population (about 3.5 million) could not get EMS within 15 minutes, compared to 7.5 % with the normal condition. This study could provide a scientific reference for city planning departments to optimize traffic under inclement weather and the site selection of emergency medical facilities.
- Preprint
(1521 KB) - Metadata XML
- BibTeX
- EndNote
Yuting Zhang et al.
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2022-218', Anonymous Referee #1, 07 Oct 2022
The study looks at the spatial accessibility of emergency medical service facilities in Beijing resolved according to weather situations, days and time-of-the-day. The authors demonstrate an empirical approach for linking spatially resolved accessibility decreases to weather situations, and manage to point out spatial inequalities on a sub-urban level. The approach, despite the reviewer having some reservations regarding certain over-simplifications along the way, shows a way forward to combine empirical data to yield insights into an under-studied topic (EMS accessibility) while shedding light on a dimension of social inequity.
General
L 28 not clear at that stage what is meant with « coverage rate ». Is it the total area covered within a 15 mins response time ? i.e. 13 % reduction refers to a km2 number ? why then « rate » (I do know that it is explained later on in the text, however the abstract should be understandable before this)?
L 45-55 : It is good that the authors give detailed insight into what parts make up EMS, especially tailored to the case study context. However, there seems to be - strictly speaking - some inconsistency in the use of the term services : One case includes the transport to the EMS facilities within the service definition (aka, the transportation service), the other case not (only treatment service
Further, please elaborate how this goes together with the definition in L 72-74 : It seems that this definition covers only the case of responders starting out from an EMS facility, getting to the scene, and transferring back to a respective facility (e.g. via ambulance). From the initial description above, the reader might think the authors will cover, however, also the case where patients transfer directly from the scene to an EMS facility (e.g. via private transportation). Hence, it might be good to explicitly mention again that this latter case will not be covered, even though described above for reasons of completeness.
L 58 : To the non-local reader, it is unclear whether 1.5 to 2 hours response time is significantly longer than normally. One would assume this, but it would be helpful to provide average response times during normal conditions for comparison.
L 78 : Reference to the 2SFCA method ?
L 74-L104 : it is good that you provide some literature revolving around the topic of study. However, this very brief listing-style does not make it very clear how those papers are related to the concrete problem statement, or not summarized into coherent areas of challenges, seeming a bit randomly aggregated.
Section 3.1:
Could you please elaborate the reason for averaging to daily speeds for the baseline constructions, since you later also look at rush-hours and non-rush-hours specifically?
In a similar line of argumentation, averaging hence-obtained speed reduction rates across all road sections within the city (L220f), seems to conceal congestion hotspots? Please elaborate why this was done and the potential limitations of this.
Also, it is not clear to me from this description, if days are simply divided in a binary manner into inclement and non-inclement weather days, irrespective of the precipitation magnitude? Please elaborate in more detail if this is the case, and what was the reasoning for and potential shortcoming of this.
Section 3.3:
How does aggregation of the population grids to 1000m in a city distort potential travel patterns? Given that the topology of the road network within Beijing is at a much higher resolution, does this aggregation not lead to a very coarse estimation of what roads are taken, and which ones not?
Section 4. Results:
L275-280: It is hard to understand to which scenarios / analysis areas the percentages belong. Do 38 and 40%, resp., refer to the city including suburban areas? And 77 and 83% refer to only the inner city (is that meant by Six Road Ring)? Please phrase it in a way that describes the area analysed better to a non-local.
L 283-288: The definition and selection of a precipitation event belongs to the methodology section.
4.1.1. When I first read that you average out the total precipitation, across the grid cells, I was sceptical whether this conceals local effects, as one might assume that certain parts of the city could hence flood more, and cause over-proportional traffic delays. Also, I do not see an analysis of total precipitation on traffic speed, which I could imagine has quite an impact (while it is certainly important how strongly it rains in a given hour, it surely also matters how long it rains for causing pluvial flooding). Please elaborate more on both those aspects. Please also use figure 3 in justifying your assumptions / method, as indeed, it seems that without the (explained) outliers, there seems to be not much of an impact on how much it rains for causing travel reductions? This may be an argument in favour of your decision; however, it is somewhat unintuitive why there is so little effect.
In general, please comment more on the relationship between precipitation and urban pluvial flooding, and limitations of looking only at precipitation data without any hydrological modelling associated with it, that would link precipitation with the actual amount of water on the streets.
Figure 4 is basically not commented and further analysed. Please elaborate more, what one can see.
Grammar / Style
In-line citations are ill-formatted (brackets around them), e.g. L 83, L 90, etc. Please format correctly. Also, some citations are CAPITALIZED.
L31: towns new sentence : Furthermore, …
L 106: Could you briefly explain the term “waterlogging” (e.g. the saturation of ground with water), as it may not be clear to every single reader what is meant by this phenomenon.
L145: Rather: “hit” by a rainstorm? They do not malevolently “attack”.
L 145-150: You already gave quite a few examples of hazardous events in the introduction; also, this example does not fit the section “Study area”. Please consider to delete it.
Section labels are wrong. For instance, after 2.1 and 2.2. on p7, we see another section 2.1 and 2.2. on p8
L261: The term “population medical accessibility index” is a little bit cumbersome to read and understand. Could you perhaps think of a simpler, more descriptive term?
Citation: https://doi.org/10.5194/nhess-2022-218-RC1 -
AC1: 'Reply on RC1', Kai Liu, 01 Dec 2022
The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2022-218/nhess-2022-218-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Kai Liu, 01 Dec 2022
-
RC2: 'Comment on nhess-2022-218', Anonymous Referee #2, 10 Oct 2022
Summary: This paper presents a study that evaluates the spatial accessibility of emergency medical services during inclement weather, including rain and snow, and measures the impact of precipitation on traffic speeds. It compares the accessibility of emergency services during inclement weather to a baseline value calculated two weeks before the event and two weeks after the event. The results highlight four days when emergency medical service accessibility particularly decreased. The study also shows that snow has a particularly large impact on emergency service accessibility. The study has the potential to provide a scientific basis for discussions with transportation and urban planners to improve access to emergency medical services, particularly in rural areas or areas with unequal conditions.
General comments:
- The study includes examples of natural hazards and the difficulty of reaching emergency services in a timely manner (L. 55- 63). Can you provide the references for these examples?
- The study presents several case studies that use different models (L. 70 -104). Could you please summarise the research gaps in this area?
- The text gives a good description of the resolution of the data used. In line 157, please define "inclement" and "normal" weather in the datasets. Is a little rain already considered bad weather?
- Some sentences are very long sentences and compromise readability:
- L. 22 – 25 ("and" is used twice in short succession)
- L. 50 – 55
- L. 74 – 79
- L. 91 – 95
- L. 117- 123
- L. 141 – 145
- L. 366 - 373
- L. 77: Please refer to the correct citation style and do not capitalize references: “Jones and Bentham…”.
- L. 100: Could you please write out the abbreviation “PF-prone” when it is first mentioned?
- L. 132: Instead of referencing the link in brackets, please refer to the correct citation style.
- L. 145: “Beijing was attacked by a rainstorm…”. Could you please paraphrase this sentence?
- L. 162: Section 2 has a wrong numbering of the subsections. Should this be 2.2.1. instead of 2.1, 2.2.2. instead of 2.2?
- L. 171: Instead of referencing the link in brackets, please use the correct citation style.
- L. 183: “third-level grade-A hospitals.” Could you please provide a brief explanation of hospital classifications that might help readers if they are not familiar with it?
- L. 255: Could you please write out the abbreviation of “OD”, when it is first mentioned?
- L. 465: Could you rephrase the phrase "we could guess"?
- Figures:
- Could you please specify which software tools you used to create the figures?
- L. 305: Figure 4 is difficult to read. Is it possible to highlight some particular days with observations?
- Discussion:
- L. 435: In the discussion, it would be good to refer to the previously mentioned studies in the introduction and draw a link: How does this work build on the previously published literature body? Where do the results align, where do they differ?
- L. 467: As next steps, you mention that future studies should consider data on “extreme precipitation” events. Are there other data analyses that can be done with the available data?
Specific comments:
- L. 24: Although it is mentioned in the Abstract, "inclement weather" is quite general. Later, in the introduction, the study refers to "rain or snow" (line 51). How much rain or snow is considered inclement weather, or is a little rain already inclement weather?
- L. 63 - 65: Since this is a very general context, could you please provide some more references?
- L. 78: Could you please name some references that use the 2SFCA method?
- L. 112 - 113: Could you please state the contribution of the study more clearly?
- L. 157: Can you give a brief description of the road network topology?
- L. 203: How many days with precipitation were included in the sample?
- L. 298: The analysis focuses on specific holidays (July 1st, September 10th). How transferable are the results of your study to other days?
- L. 254: “population medical accessibility index”. The term can be a little difficult to understand. Can you briefly explain the term in more detail?
Technical corrections:
- L. 31: “towns with lower baseline EMS accessibility are more vulnerable to inclement weather. Furthermore,”.
- L. 53: For quotations in continuous text, please insert a space in between the text and the reference: “The efficiency of emergency services is highly vulnerable to inclement weather conditions[...], and sometimes block roads completely (Agarwal et al., 2006;…”
- L. 152: For quotations in continuous text, please insert a space in between: “Andersson and Stålhult (2014) used network analysis”
- L. 188: How about phrasing the sentence: “The data records present the population size" or “The data records depict the population size…”?
- L. 192: How about phrasing the sentence: “Figure 2 presents” or “Figure 2 illustrates”?
- L. 315: Is it “In which the 15-min EMS coverage rate reduced by …”?
- L. 319: “…which led to a significant reduction in overall EMS coverage…”
- L. 418: Here, should it be “within the Sixth Ring Road extent”? Later, in line 363 and in line 365, the text refers to “within the Sixth Ring road”.
- L. 428: “…were almost no regions where the population medical accessibility index decreased.”
Citation: https://doi.org/10.5194/nhess-2022-218-RC2 -
AC2: 'Reply on RC2', Kai Liu, 01 Dec 2022
The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2022-218/nhess-2022-218-AC2-supplement.pdf
-
RC3: 'Comment on nhess-2022-218', Anonymous Referee #3, 14 Oct 2022
This empirical study investigates the impact of inclement weather on the time emergency medical response (EMS) time intervals for the city of Beijing. It is broken into two stages. Firstly, to explore the impact of inclement weather (i.e., precipitation) on traffic and EMS accessibility to come with the worst-case scenarios of the year 2029 (i.e., days including different times per day). Secondly, to evaluate EMS accessibility under the identified worst-case scenario and evaluate the distribution of EMS with particular focus on the difference in population and road network distribution between urban and suburban areas. The study can be useful to identify the scenarios needing improvement to ensure more fair access to EMSs for populated cities. The paper is generally well-written and easy to read but can be improve in terms of clarity.
The abstract. It seems to overlook a key impact that seems important from the results: The day of snowfall seems to have more significant impact that the days of rainfall (among the worst-case scenario considered). Can the authors add a mention about this fact somewhere in the abstract.
The introduction. In Line 107, Please be specific on what is meant by “The latter”. It can be more effectively used to also introduce to lay reader some common terms that will be occurring later, such as “coverage area” and “waterlogging”.
What platform was used to “Combining the topology road network with medical facility locations and the distribution of the population, we could further analyse the spatial accessibility to EMSs.” Was this work GIS based? What was the tool employed?
Line 162: the sub-section numbering here down to line 189 is missing a third digit. I suppose it should be 2.2.1, 2.2.2, 2.2.3 and 2.2.4.
Methodology. As in any study, there should some assumption made by the authors ands aspects that were not addressed. A mention of these would be useful.
Line 232: Is there any citation that you can use to justify the choice of the 15-minute arrival time?
Line 255: please define OD. Does it stand for Origin-Destination?
Line 258-259: More discussions on the calculation cost is welcome so one can justify the rationale behind increasing the resolution from 100 and 1000 m. How much would this impact the predictions vs. reduce the cost for the analysis?
Line 265: there is always a mention of a grid. Should this be meaning that the analysis was GIS based?
Line 306: Figure 4 deserves a discussion as such in 4.1.1 before moving to 4.1.2 and quoting it there. I guess it was used to support further the choice for the days considered as worst-case scenario.
Line 331: “The results demonstrate …” what results? Any figures I should be looking at? Or, from which equation? Are you talking about the “coverage rates”. Please specify.
Line 359: The clarity of the sub-figures in Figure 7 can be significantly improved. Same for Figure 8.
Overall, the snowfall seems to have the greatest impact and it useful to highlight in key locations including the abstract and conclusions. This could be hinting at the fact that such a study is more of relevance to cities affected by snowfall.
Citation: https://doi.org/10.5194/nhess-2022-218-RC3 -
AC3: 'Reply on RC3', Kai Liu, 01 Dec 2022
The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2022-218/nhess-2022-218-AC3-supplement.pdf
-
AC3: 'Reply on RC3', Kai Liu, 01 Dec 2022
Yuting Zhang et al.
Yuting Zhang et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 484 | 120 | 24 | 628 | 7 | 9 |
- HTML: 484
- PDF: 120
- XML: 24
- Total: 628
- BibTeX: 7
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1