On the Use of a Bike-Sharing System in Extreme Weather Events: The Case of Porto Alegre, Rio Grande do Sul, Brazil

Abstract

This article aims to analyze the use of a bike-sharing system (BSS) during the flooding event caused by extreme rainfall that hit the municipality of Porto Alegre, Brazil, in May 2024. Public transport services were interrupted, prompting an investigation into the resilience of the BSS during the crisis. Considering data from the Tembici BSS company, a set of approximately 400,000 trips made between 104 stations in the municipality of Porto Alegre from January to May 2024 were analyzed. Daily rainfall data from the National Institute of Meteorology (INMET) were compared with the daily trip flow to identify the travel flow patterns on the days most affected by the flooding. The results indicate an abrupt drop in shared bicycle use during May 2024, but 7600 trips were recorded despite the crisis. Regarding the travel pattern between 1 May and 10 May, most trips were still for recreational purposes (73%), while trips for work and study accounted for 22% of the total, and only 5% were for delivery services. Overall, the resilience of the BSS during the extreme climate event in question points to the continuation of practical daily activities, although with more significant effects on economic-related activities and lesser effects on leisure-related activities.

1. Introduction

Climate change can be considered on global, regional, and local scales. The effects of climate change, or the climate crisis, contribute to the occurrence of extreme events, such as extreme rainfall, and their repercussions, including flooding, landslides, heat waves or cold spells, droughts, wildfires, and forest fires, among others [1,2,3].

Urban areas, characterized by extensive impermeabilization, verticalization, intense modal flows, and pollutant emissions contribute to the formation of heat islands, which significantly amplify the occurrence of extreme events [4,5]. In Brazil, the incidence of these events has become frequent, as observed in 2023, when 12 extreme climatic events were recorded, nine of which were considered unusual and two unprecedented. One of these unprecedented events was an extratropical cyclone, with torrential rains, floods, and severe impacts on cities in the state of Rio Grande do Sul—a precursor to the disaster that would strike the state in 2024 [6,7,8].

News such as “Flood affected nearly 12% of Porto Alegre’s population”, “Porto Alegre under landslide alert” [9], and “Return of rain causes flooding and road blockages in Porto Alegre” dominated the media, highlighting the extent of damage caused by the floods to the city’s population in May 2024 [9,10,11]. These events also emphasize the challenges of addressing and fulfilling the commitments of the global environmental agenda about resilience, mitigation, and adaptation to climate change for local governments. Climate resilience requires comprehensive governance systems and inclusive urban development models that can provide sustainable adaptation measures and means to reduce the damage caused by the climate crisis [12,13].

Various sectors are affected by extreme climate events, including the economy, health, education, and urban mobility. Urban mobility mainly facilitates people’s movement and flow for various purposes through transportation modes. Understanding how these movements occur is a challenge, especially in the context of climate crises, when there is a reduction in service availability and difficulties in quantifying displacements [14,15].

To mitigate the impacts of climate change, urban mobility provides sustainable alternatives for commuting within cities. However, significant challenges remain, including the entrenched culture of individual motorized transport and the lack of infrastructure to support citizens opting for active transport modes. Among these modes, the bicycle stands out. While it already serves as the primary mode of transportation for a substantial portion of the population in small and medium-sized Brazilian cities, its role is increasingly gaining prominence in large urban centers [16,17].

The urban road network must be rethought in order to effectively achieve the perspective of resilient cities, ensuring a high level of quality of life and well-being. This is necessary in view of the significant emission of greenhouse gases (GHGs), such as carbon dioxide resulting from fuel combustion, primarily due to the increase in the number of vehicles in cities since the late 20th century [18].

In addition to reducing greenhouse gas emissions, cycling can offer health and quality-of-life improvements, as well as reduced commuting times in urban centers experiencing road saturation caused by the widespread use of individual transport. However, heavy motor vehicle traffic in cities creates unsafe conditions for cyclists, who have a high risk of accidents [19,20].

Among the most modern alternatives for active mobility is using bicycles through sharing systems that allow for multimodal integration and travel within the compact areas of large urban centers [21,22]. However, residents of peripheral areas face challenges due to the distance between home and work, compounded by the lack of cycling infrastructure [23]. Compact cities offer population density and accessibility, with shorter distances between services and residences, and promote environmentally friendly transport options such as active mobility [24]. In these areas, micromobility has gained importance, with the use of lightweight vehicles and the potential for multimodal integration [25]. The use of shared bicycles has become popular as a solution to the lack of urban planning, fostering sustainable mobility in large urban centers [26].

A bike-sharing system (BSS) is defined as a network of stations implemented by a city where bicycles are made available to users. In Brazil, the startup Tembici is one of the companies providing this service, operating in ten Brazilian cities. The company categorizes users based on the purpose of their trip: recreational, utilitarian (work and study), or service-related [27,28].

These systems are inherently private and have become increasingly popular in recent years due to the spread of digital platforms that facilitate more practical and faster integration between users and applications. They also allow users and service providers to monitor, in real time, the availability of bicycles and the mechanisms of retrieving them from the stations [29,30].

In this complex scenario, this article aims to investigate the impacts of heavy rainfall and flooding on urban mobility in Porto Alegre, Brazil, and the resilient behavior of daily shared bicycle use as an active mobility alternative. For this, a data-driven approach is adopted to support our discussions and insights. More specifically, the two objectives of this article are the following:

• To identify and quantify daily bicycle trips during periods of flooding and heavy rainfall, as well as the demographic characteristics of the affected areas.
• To analyze the relationship between rainfall indices and flooding, using shared bicycles based on user profiles, during extreme events and their repercussions.

The remainder of this article is organized as follows. Section 2 presents some related works that influenced our investigation. Section 3 describes the materials and methods in this article, supporting all contributions. Then, Section 4 presents a data-driven analysis of bike trips in the defined scenario. Discussions and analysis are described in Section 5, followed by Conclusions and References.

2. Related Works

Urban flooding has become a critical challenge in cities worldwide, driven by the increasing frequency and intensity of extreme weather events [31,32]. Recent studies have emphasized how climate change exacerbates urban flooding, leading to severe social, economic, and infrastructural disruptions. In densely populated urban areas, flooding often damages housing, transportation systems, and public utilities, creating cascading effects that disproportionately impact vulnerable populations [33,34].

There has been growing interest in climate-resilient urban mobility systems in response to these challenges. Research has explored how transportation infrastructure and services can adapt to extreme weather conditions, particularly flooding. Previous works have shown that traditional urban transportation systems, such as buses and trains, are often highly vulnerable to waterlogging and road blockages, making mitigation and rescue more challenging [35,36].

Technology-driven smart mobility solutions have further contributed to the resilience of urban transportation. Digital platforms, real-time monitoring systems, and data-driven decision-making processes have facilitated adaptive responses to mobility disruptions [37]. In parallel, there has been a significant increase in active transport modes, such as cycling, as sustainable alternatives to traditional forms of commuting [38,39]. However, there are few studies linking the topic of extreme climatic events with the use of active transport modes such as bicycles and electric scooters. Too few works have related the mobility mode used and the spatial conditions of commuting, concerning urban structures and the impact of inequality on the population’s mobility conditions, to the effects of climatic events on the choice of transport mode [40].

In recent years, interest in the relationship between urban mobility and extreme weather events has grown. An example of this is the study of urban mobility and resilience from the perspective of social dynamics and institutional factors, expanding the discussion beyond technology, engineering, and urban infrastructure [41]. Within this scope, there is an emerging interconnection between climatic factors, bike-sharing systems, and the social dynamics of the population, as well as how external factors influence the use of these systems and promote improvements in urban center mobility [42,43]. However, significant gaps remain in understanding how these phenomena interact, particularly regarding the use of bike-sharing systems (BSSs) [44].

Existing studies often examine urban flooding or smart mobility in isolation, without considering how extreme weather conditions influence the choice and operation of active transport modes. Additionally, there is a lack of quantitative analyses that explore the relationship between user demographics, flood severity, and bike-sharing usage patterns during such events. Based on this, the following research gaps can be identified:

• A lack of integrated studies on active mobility and extreme weather events: Most research examines urban flooding or smart mobility separately, without exploring how adverse weather conditions affect the choice and operation of active transport modes, such as bike-sharing systems.
• Scarcity of quantitative analyses on the impact of flooding on bike-sharing usage: Few studies correlate flood severity, user profiles, and variations in BSS usage patterns during extreme events.
• Limited use of high-frequency databases: Most studies rely on aggregated or long-term data, making it difficult to capture daily behavioral changes in urban mobility, especially during crises, when conventional public transportation becomes unavailable.

Given these gaps, this study aims to contribute to the understanding of urban mobility in extreme weather scenarios, with a particular focus on bike-sharing systems in Porto Alegre. The main motivations for this research are as follows:

• Understanding the impacts of heavy rainfall and flooding on urban mobility: Investigating how extreme weather events affect bike-sharing users’ behavior can help inform public policies aimed at enhancing urban resilience.
• Providing evidence for the adaptation of urban infrastructure and transportation systems: Analyzing BSS usage patterns during extreme weather conditions can support the planning of cities better equipped to handle environmental challenges.
• Exploring the potential of non-traditional databases: The use of high-frequency data can offer a more detailed view of rapid changes in urban mobility, aiding in the development of effective disaster response strategies and promoting active transportation as a viable alternative in emergency situations.

3. Materials and Methods

Transportation is essential for urban mobility, allowing people to travel from their homes to various destinations within cities, whether for work, study, leisure, healthcare, or shopping. Understanding these movements is not a simple task. Measuring transportation flows requires specific methodologies, such as an origin–destination survey, which aims to identify the different modes individuals use to navigate the city [45].

Official surveys and data collection efforts, such as the Population Census and the National Health Survey, also gather information on the transportation modes used by the population. However, these studies typically record only the primary mode used during the trip, disregarding complementary segments between the origin and destination. As a result, active transportation modes, such as bicycles, tend to be under-reported. In this context, non-official statistics, derived from data shared by companies and collaborative platforms, help bridge this gap by providing more detailed insights into the use of these modes, as seen in the case of shared bicycles [46].

Tembici, a startup that operates a bike-sharing system in Brazil, provides data that identify trips made in Brazilian urban centers where the service is available. In the case of Porto Alegre, the Tembici project is known as BikePoa. The availability of origin and destination data from bike stations sheds light on short to long-distance daily trips, which are not captured by other official surveys [47,48].

This data source compiles information from bike-sharing system users, usage time, and georeferenced trip data from stations that the company’s operational system collects.

Each row in the database represents a trip, including information about the origin and destination stations, the geographic coordinates, the date, and the duration in minutes. The files are stored monthly in .csv format and were provided by the company under a usage license [49].

The columns contain information about the trip profile, type of bicycle, origin and destination stations, geographic coordinates of the origin and destination stations, date, and trip duration in minutes. The variables contained in this database are described in

Table 1. Description of variables in the Tembici database. Adapted from [49].

Variable	Description
Trip Duration	Total trip duration in minutes
Trip Date	Date when the trip occurred
Start Time	Time the trip started
End Time	Time the trip ended
Origin Station	Name of the station where the trip began
Destination Station	Name of the station where the trip ended
User Profile	Classification of the trip: recreational, utilitarian, or service
Type of Bike	Classification of type: mechanical or electric

.

Although this study focuses on the climate emergency period that occurred in May 2024, data are available from 2018 up to the most recent update. However, to analyze mobility patterns during the floods in Porto Alegre, data from May 2023 and 2024 are compared, with the most relevant data being those that indicate the quantity, duration, and movements based on location information such as the latitude and longitude of departure and arrival points. Hence, the coordinates of the BikePoa bike stations are considered [49].

In addition to examining the number of trips made, this study also assesses the distribution of shared bicycle usage based on the type and profile of users. This approach helps to identify which groups of cyclists continued using the BSS service, as well as to analyze the average daily usage time, the most in-demand days of the week, and the most frequent travel times. Furthermore, the stations are ranked according to their usage, highlighting the 15 most utilized ones, which are later mapped to represent the areas of the city where shared bicycle traffic was highest during the flood period.

The results are presented in line charts, histograms, and thematic maps. This study conducted the descriptive analysis using the Python programming language in Google Colaboratory (version 3.13.0) and R (version 4.4.3) [50].

3.1. Porto Alegre City

In Porto Alegre, the 2022 Demographic Census recorded a population of 1.3 million inhabitants, making it the second most populous capital in the South Region of Brazil. However, the same survey indicated a demographic decrease of 5.4% in the city. Porto Alegre shows an increase in population aging, surpassing the national average [51]. The city has a diversified economy and a high Human Development Index (HDI) of 0.840 [52].

Facing several environmental challenges, Porto Alegre has recently experienced serious problems, especially floods resulting from intense rains. The geographic location of the municipality, bordered by Guaíba Lake to the west and south and the Jacuí River to the north, contributes to these issues, which have recently affected approximately 12% of the population [53,54]. Figure 1 depicts the geographical area of the city.

Although the city has a public transportation network, including a subway system and urban bus lines, the crisis caused by the floods resulted in the interruption of these services. Therefore, there was an institutional effort, both public and private, to mitigate the crisis. However, in September 2024, the urban mobility in Porto Alegre was still in the process of infrastructure recovery, with completion expected by December 2024 [56,57,58].

3.2. Meteorological Data

Rainfall is one of the climatic elements that most influences and exacerbates social and environmental issues for populations. Its analysis on an urban scale, especially during extreme weather events, challenges infrastructure, transportation modes, terrain, and, most importantly, housing, health, and transportation conditions for populations affected in various ways.

In the context of the climate crisis, extreme events caused by rainfall must be analyzed from a historical and spatiotemporal perspective spanning over 30 years, as recommended by the World Meteorological Organization (WMO) [1,59].

For this study, the frequency analysis considered rainfall data from 1993 to 2024 for Porto Alegre. The data were sourced from the meteorological database of the National Institute of Meteorology (INMET). To estimate extreme rainfall, daily data from the historical period—1 January 1993 to 31 May 2024—was analyzed at the 95th percentile (extreme upper values for accumulated rainfall in a given year), indicating that rainfall above 41.5 mm qualifies as extreme [60].

Alongside daily, monthly, and annual accumulated precipitation, the average elevation of flood points in Porto Alegre was analyzed for May 2024. The information on average flood elevations for May is part of a map repository that documents the tragedy in Rio Grande do Sul, providing data by measurement areas that differ from the territorial divisions of the Brazilian Institute of Geography and Statistics (IBGE). The repository was organized by researchers from the Hydraulic Research Institute, the Institute of Geosciences, and the Faculty of Architecture (FAUFRGS), as well as volunteers and collaborators from other units and institutions [61].

Finally, data on the Guaíba Lake water level were used, available at two points in the city of Porto Alegre, Cais Mauá C6 and Usina do Gasômetro, compiled by the National Water and Basic Sanitation Agency of Brazil (ANA). This information includes the daily lake level and the safety threshold of 3.60 m, which, when reached, causes flooding in areas near the lake. These data are presented alongside shared bicycle usage data to support the descriptive analysis of the influence of flooding on bicycle usage [62].

3.3. Demographic Data

The 2022 Demographic Census, produced by the Brazilian Institute of Geography and Statistics (IBGE), was used as the cartographic basis for generating maps and distributing the enumerated population into neighborhoods [51]. Based on census information on the resident population and the area of the neighborhoods, the indicator inhabitants per square kilometer was used to express the degree of population concentration in a given area. The variables used are described in

Table 2. Variables considered for calculating the population density indicator by neighborhood. Adapted from [51].

Variable	Description
CDMUN	Municipality Code
NMMUN	Municipality Name
CDBAIRRO	Neighborhood Code
NMBAIRRO	Neighborhood Name
V0001	Total Residents
AREA km2	Area of the neighborhood in km2
Geometry	Geospatial information

.

To assess the density of households and services in the neighborhoods and evaluate the impact of the floods on social life in the municipality of Porto Alegre, data from the National Address Database for Statistical Purposes (CNEFE) of the 2022 Demographic Census were used, specifically focusing on the geographic distribution of address points [51]. Four thematic maps of the city of Porto Alegre are presented, based on the type of address, showing the distribution of households, educational institutions, healthcare facilities, and other establishments [51]. The variables used are presented in

Table 3. Variables considered for calculating the density of households and services by postal address code Adapted. Adapted from [51].

Variable	Description
CEP	Postal address code
TOTDP	Total households
TOTEENSINO	Total educational establishments
TOTESAUDE	Total health establishments
TOTEOF	Total establishments for other purposes

.

3.4. Data Processing

The presence of outlier data in research is common, making it necessary to identify such values to determine an appropriate methodological strategy for data adjustment, thereby avoiding inaccuracies or bias in the analyses [63]. Based on the research context, the removal method was adopted. The selection of the variable imputation method was guided by the proportion of outlier values, which in this case ranged from 5% to 15%. Within this range, simple imputation is still recommended.

In the case of the Tembici database, patterns of outliers were observed in the variable trip duration. Approximately 6% of the trip duration data showed anomalous values, with travel times significantly shorter or longer than the typical range, exceeding several days or even weeks.

A filtering function was applied to remove values above 240 min and below 5 min to address this. This threshold was determined based on the plan with the longest allowed usage time for shared bicycles for the upper limit and by accounting for potential incidents during bike withdrawals at stations for the lower limit. As a result, data from 100,000 trips from the BikePoa Project were analyzed from May 2023 and 360,000 trips were analyzed from January to May 2024.

Regarding the other variables used to obtain the results, no inconsistencies were identified in the data, as they are automatically generated by the system and are therefore not influenced by external factors, such as difficulties in retrieving bicycles from stations or mechanical issues after use [37,64].

4. Results

As observed, within the analyzed period the service is more utilized in January, which corresponds to the summer season and school holidays, with approximately 100,000 trips recorded. From February to April, the number of trips stabilizes, followed by a sharp decline in May, coinciding with the flood crisis, caused by heavy rains that impacted Porto Alegre.

Figure 2 shows the number of shared bicycle trips made during the first five months of 2024.

According to Figure 2, a persistent decline can be observed starting at the end of April, marking the onset of heavy rainfall that led to flooding in Porto Alegre. This trend continues in May, with zero usage recorded on eleven days of the month.

In this context, Figure 3 provides detailed information on May, the critical period of the crisis, during which only 7600 trips were recorded in 2024, while in 2023, more than 100,000 trips were registered, further highlighting the influence of precipitation and its consequent impact on bicycle usage.

The accumulated daily rainfall for May 2024 reveals the first period of extreme rainfall occurring specifically on 1 May 2024. Subsequently, a decrease in accumulated daily rainfall is observed between 2 May and 9 May 2024, followed by a second period of extreme rainfall lasting for two days (10 May and 11 May 2024). Another reduction in rainfall is noted before a new extreme rainfall event occurs, with over 102 mm of rain recorded on 23 May 2024.

During the second and third periods identified as extreme rainfall events (10–11 May 2024, and 23 May 2024), the use of bicycles was almost nonexistent, with only 37 trips recorded on 10 May 2024, and no trips at all on 11 May and 23 May 2024. The decrease in the use of shared bicycles can be partially explained by the state of emergency declared by the government starting on 29 April 2024. However, this change can also be attributed to a perceived climate crisis event experienced by the residents of Porto Alegre, who faced the impacts of extreme rainfall. This situation led many to stop using the bike-sharing system, and the search for alternative public transportation, such as buses and subways, combined with the effects of rain and flooding, further intensified the decline in the use of shared bicycles.

Although rainfall was recorded for approximately ten days in May 2023, leading to a sharp decline in usage, it was not enough to render the system unusable. The data suggest that the absence of trips in May 2024 is due to the influence of high rainfall and flooding, reducing shared bicycle use. This reveals the vulnerability of the transportation system to compound extreme weather events, as found in the study in [65].

Another metric that helps clarify usage patterns during the flooding period in Porto Alegre is the weekly trip distribution, as shown in Figure 4. In May 2023, a usage pattern is observed with high utilization on Mondays, which may indicate a utility-based travel profile, and the lowest value on Saturdays.

This behavior changes in 2024, with a decline in usage frequency on weekdays, with Wednesday having the lowest number of trips, around 600. On the other hand, weekends see the highest usage, with Saturday surpassing 1600 trips, which may indicate a recreational travel profile.

Table 4. Comparison between the month of May in the years 2023 and 2024 [49].

Information	May 2023	May 2024
Average trip duration in minutes	34 min	33 min
Day of the week with the highest number of trips	Monday	Saturday
Station with the most trips started	Skate Park	Rótula Gasômetro
Station with the most trips completed	Skate Park	Rótula Gasômetro
Most frequent travel profile	Utility (40%) Service (7%) Recreational (53%)	Utility (22%) Service (5%) Recreational (73%)
Most frequent type of bicycle	Electrical (3%) Mechanical (97%)	Electrical (3%) Mechanical (97%)

details other characteristics of Tembici trips in the BikePoa Project for May 2023 and May 2024. This comparison between a typical rainy month in the region (May 2023) and an atypical May in terms of rainfall and its repercussions, such as floods (May 2024), provides important insights.

It is observed that, in terms of trip duration, the emergency situation in 2024 does not appear to have affected the average trip duration, which remained around half an hour, or the type of bicycle used, whether mechanical or electric.

However, considering the average trip duration for each day in May 2023 and May 2024, as shown in Figure 5, it is observed that in 2023, shared bicycle users had a peak usage on weekends, exceeding 40 min, while maintaining lower average usage above 25 min.

During the first ten days of the month of May 2024, the average trip duration remained quite stable, with a decline of 33% between the 1st and the daily average up to the 10th, fluctuating between 40 and 30 min. Starting on the 11th, there were 14 days with zero average values due to the floods, as the service was not in use. In the subsequent days, however, peaks in the average usage time were observed, such as on the 15th, when it reached around 130 min. This spike can be explained by the low number of users on that day, with the usage time concentrated on a few individuals.

Flood events had distinct impacts on the profiles of shared bicycle users, as illustrated in Figure 6. The figure highlights shifts in usage patterns between the observed periods in 2023 and 2024, notably a 45% decline in utility trips and a 38% increase in recreational trips. This shift can be attributed to the decrease in work- and study-related trips, which are characteristic of utility users, following the state of public emergency declared due to the floods.

Meanwhile, the service user category also saw a 28% decrease, yet recorded the longest average trip duration, approximately 140 min. In contrast, utility and recreational users averaged 10 and 20 min, respectively, during the climate crisis event. Of significance is the high usage near shopping centers, driven by service-oriented users, particularly delivery riders for the company Ifood, representing 5% of users. It is worth noting that among all users in the services profile, 50% used electric bicycles. However, this type of bicycle accounts for only 3% of the total available in the BikePoa project, while mechanical bicycles make up the remaining 97%.

Therefore, heavy rainfall combined with soil impermeability and the rising level of Lake Guaíba, worsened by the malfunctioning of the water drainage pumps, led the city of Porto Alegre to experience the most severe flooding in its history. This reinforces the premise that heavy rainfall alone was not enough to prevent the use of shared bicycles, but rather it was due to the occurrence of flooding. As shown in Figure 7, even with heavy rainfall during the first ten days of the month, bicycle trips still took place. However, on 3 May, when the water level of Guaíba Lake reached the flood threshold, there was a sharp decline in trips, dropping to nearly zero for almost the entire month. This situation persisted until 1 June, when the lake returned to a safe level.

Lake Guaíba rose to 5.37 m above its normal level, and the floods affected about 12% of the city’s population. Along the Jacuí River, which traverses the city, 46 neighborhoods were recorded as having flood points during the climate crisis event.

The neighborhoods most affected by the floods were concentrated in the north zone, Sarandi, Santa Rosa de Lima, and Humaitá, and in the central region, Menino Deus, Cidade Baixa, and Centro Histórico, as shown in Figure 8.

The flooding began in the lower areas near the Guaíba Lake waterfront and extended throughout the city along the Jacuí River basin. Other neighborhoods can also be highlighted regarding the relationship between flooding height and the presence of shared bicycle stations, such as Praia Belas, Cristal, and Ipanema, all located around Lake Guaíba. Only the Cascata neighborhood was unaffected by the floods due to its topography, characterized by hills, which distinguishes it from many other neighborhoods in the city [49,66].

As illustrated in Figure 9, population density in each neighborhood was also a key factor in the impact of the floods, as areas with higher density were more severely affected. Among the ten neighborhoods with the highest population density, five experienced flooding in May 2024: Bom Fim, Cidade Baixa, Independência, Centro Histórico, and Menino Deus [51].

The neighborhoods of Bom Fim, Cidade Baixa, and Independência have resident populations of around 10,000 inhabitants. However, their areas are relatively small, each being less than 1 km². In contrast, the Menino Deus neighborhood has a larger area of 2.3 km² and a population that is nearly double that of the previously mentioned neighborhoods, totaling 27,961 inhabitants [51].

The distribution of shared bicycle stations highlights the limitation of service expansion to other neighborhoods in the city, with a higher density of stations in the western portion of the city and scattered points in the central region (Partenon, Petrópolis, and Jardim Botânico) and neighborhoods along the Guaíba Lake waterfront (Cristal, Vila Assunção, Menino-Deus, Centro Histórico, and Ipanema).

Population density, along with the presence of households and the distribution of shared bike stations—particularly those most used (Skate Park and Rótula do Gasômetro) for commuting in May 2023 and 2024—aligns with the density of available health services, educational facilities, and other types of establishments, especially near the Guaíba Lake, as shown in Figure 10.

It can be observed, therefore, that the mentioned areas embody a compact city model. This is because, as the expansion into peripheral neighborhoods occurs, there is a reduction in the availability of health and educational services, and most notably an absence of shared bike stations, which, in theory, would improve urban mobility.

In May 2023, according to Figure 11 the busiest station was Skate Park, recording over 4700 trips, likely due to the presence of recreational facilities in the area. However, in the same period of 2024, the main station shifted to Rótula do Gasômetro, with only 261 trips, marking a steep decline in a region that offers both services and public recreational facilities.

The spatial distribution of the most used stations in May showed that bicycles were utilized in areas with high flood levels, where most trips started and ended at the same station. On average, 20 trips occurred during May, primarily in the central areas of the city, which host a concentration of services, with a strong presence in the neighborhoods of Menino Deus, Cidade Baixa, and Centro Histórico. Conversely, the most used stations were located along the Guaíba Lake waterfront, with the Rótula do Gasômetro station being the most popular. This station is surrounded by recreational areas like the Iberê Camargo Foundation and Parque Marinha do Brasil.

Both stations are located near the shores of Guaíba Lake, which helps explain why, despite the floods, bicycle trips in 2024 were concentrated in the early days of the month, before the critical flood level was reached. It is common for the station with the highest number of trip origins to also be the most frequent destination, as observed in BikePoa. As observed in Figure 12, where the two main origin stations are also the ones that attract the most destination travel flows.

The Rótula do Gasômetro station, despite its proximity to Lake Guaíba, maintained trips until 5 May, partly made by users of shared bicycles commuting to public services and agencies in the surrounding area of the station.

Although there was a sharp decline in the number of trips, they still followed a similar directional pattern, with trips primarily in the western part of the city, near the lake. This could explain the shift in the most used station from Skate Park in 2023 to Rótula do Gasômetro in 2024, possibly due to flooding in the park’s recreational area during the first days of the month.

5. Discussion

This article has addressed the use of trip data of a soft mobility service when analyzing some urban transportation patterns during a critical environmental event, highlighting how these types of data can be valuable for insightful analysis, particularly in a developing country. In summary, this article adds theoretical and methodological contributions. Regarding the theoretical field, there are few studies available in the literature that discuss the resilience of active transportation modes in climate emergency events in cities in developing and underdeveloped countries. In such cases, structural failures in urban planning are common, especially concerning bicycle use, an individual mode of transport characterized by the vulnerability of its riders even under normal conditions, given the absence or precariousness of bike lanes, cycling infrastructure, and competition for road space with other users. Understanding the response of bike-sharing systems in synergy with a fragmented urban and road infrastructure contributes to the theoretical development of resilience in the face of climate emergency events similar to the one analyzed in this article.

Regarding methodological contributions, databases that provide the origin and destination of trips for any mode of transportation are rarely available. Unlike in developed countries, where bike-sharing systems are more widespread, Brazil has few studies on urban mobility via bicycles. This is due, on the one hand, to the limited adoption of such systems and, on the other hand, to the need to establish scientific partnership agreements for accessing these data, as was possible in this study. The ability to correlate travel flow information within a small geographic area—such as neighborhoods—while considering demographic variables as background enabled a detailed and up-to-date analysis of travel flows within a scenario of disruption caused by an atypical rainfall pattern in a Brazilian capital. Understanding the interaction between travel flows and the urban context represents a methodological contribution both to academia and, to some extent, to urban public policies.

The article demonstrated that, in the case of a bicycle-sharing system in the Brazilian municipality of Porto Alegre, the service is geographically concentrated, despite the variability in population density and service distribution across the territory, which could encourage its expansion. Investments in this business model, alongside urban infrastructure improvements to enhance cyclist safety—ensuring that cycling is perceived as an enjoyable experience that reduces financial and time costs—can be a key factor in expanding the system. Additionally, the potential health and environmental benefits further support its adoption by the population.

As already mentioned, climate change has repercussions on different scales, with the repercussions at the city level being the most emblematic and rapidly impacting people’s lives. In light of this, the construction and promotion of sustainable and resilient cities require interaction between international policies, national protection and civil defense measures, and urban mobility planning—specifically in terms of mobility infrastructure.

The scenario analyzed in the city of Porto Alegre indicates a growing trend in the impacts of extreme climate events and their repercussions on the lives of the population, similar to what has been observed in other Brazilian urban centers, whether due to high rainfall indices or the occurrence of severe droughts. Furthermore, it is evident that, beyond natural events, there were governance and crisis management issues on the part of public authorities, which neglected the maintenance of pumps in the flood protection system. This failure directly affected the residents of neighborhoods such as Cidade Baixa, Menino Deus, and Azenha [67].

The social function of the city involves the provision and realization of human dignity through the availability of basic services such as housing, sanitation, education, leisure, security, and culture, among others, as well as transportation—guarantees that are monitored by law. However, these are not resilient in cities, especially in terms of providing quality of life to their populations.

These extreme events, such as extreme rainfall and flooding, generally have a more substantial impact on socially vulnerable populations. In this case, the neighborhoods of Sarandi and Humaitá exemplify how extreme climate events reflect inequality and heterogeneity in urban spaces, highlighting the evidence of climate injustice. This statement is supported by the fact that 5% of users, identified as delivery service workers, continued using shared bicycle systems despite the floods and high rainfall indices to sustain their financial income.

The analysis of the relationship between rainfall indices, flooding, and the use of shared bicycles, based on user profiles during the climate crisis event, indicates that user profiles are also linked to the most frequently used stations.

However, the most used stations during this period were in areas with parks and recreational spaces, even in the vicinity of Guaíba Lake and areas with high flood levels. Recreational users accounted for 73% of the usage. Although they represent the majority of trips, this category may include users registered under the recreational mode. However, given the circumstances and observed short-distance and short-duration trips, they likely used bicycles for other purposes.

Users whose profiles fall outside the categories of service and recreation are classified as utilitarian and accounted for 22% of shared bicycle usage during the period. Most of these trips occurred near areas with access to mass public transportation, such as metro stations and bus terminals. Faced with the challenges posed by the crisis, bicycles served as an alternative for this group; however, their usage was concentrated only during the initial phase of the crisis and was completely disrupted during the most critical period.

Moreover, our findings confirm the results of Zheng et al. [68], which identified that the use of shared bicycles is more popular among young and low-income populations. The evidence on user profiles, especially the predominance of recreational users (71%) during extreme events and their repercussions, such as floods and inundations, indicates a high degree of resilience in the bike-sharing system. This is supported by Markolf et al. [69] and Ji, T. et al. [65], who showed that the system was able to maintain its functions even during periods when extreme events were occurring.

The authors indicated that the resilience of the transportation system encompasses not only its ability to prevent malfunctions due to interference, but also its capacity to avoid, adapt to, and mitigate the impact of catastrophic local events or complete system failure.

However, we can consider that the bike-sharing system is resilient in parts, especially regarding its adaptability and the speed with which it returns to operation. The general resilience of the system is not confirmed, as the mitigation of impacts caused by extreme events was not addressed, and such an objective would only be achievable if it incorporated structural planning in close coordination with public space management and its administrators.

This research provides insights into new perspectives for addressing the relationship between population segments and the environment in urban contexts. Future studies could extend the discussed approach to include risk factors associated with extreme climate events, employing more refined quantitative data analysis techniques. Moreover, analyses regarding the differences and better conditions of resilience and sustainability between different bicycles for micromobility services might be considered. The literature has already indicated that during the production phase, smart bicycles generate approximately three times the amount of greenhouse gas (GHG) emissions compared to smart docked bicycles per kilometer traveled. Thus, considering smart and resilient cities also means thinking about the entire process.

Finally, in times of crisis and climate injustice, it is urgent to consider the perception of affected people, as well as mitigation and response solutions from the perspective of sustainable urban mobility, which can improve governance systems during climate crisis events.

6. Conclusions

By promoting the use of shared bicycles as a sustainable mode of transportation with partial resilience to periods of extreme climate events, governance models and private sectors can expand the availability of these services to other urban centers, thereby making access to a low-cost and shared mode of transportation more inclusive.

The relationship between urban mobility and the environment is intrinsic and highly significant to society. Thus, the ability to analyze the impacts of extreme climate events on the daily lives of the population and explore alternative mitigation measures represents a significant step forward in addressing crises, as exemplified by the shared bicycle systems in the city of Porto Alegre.