The Impact of Weather on Shared Bikes

Abstract

This article explores the impact of weather and environment on shared bicycles. Using a random forest model combined with explanatory machine learning methods, the relationship, threshold effect, and interaction effect between weather factors and the transfer volume of shared bicycles at subway stations are analyzed. Research has shown that using the RF+IML method to study the impact of weather variables on shared bicycle transfer volume is feasible. There is a significant nonlinear relationship between various weather factors and shared bicycle transfers. Temperature, humidity, and rainfall have specific activation and threshold effects on the number of shared bicycle transfers. When humidity is below 60%, the variation in transfer volume remains relatively stable; however, once it exceeds 60%, the transfer volume drops sharply. When the temperature exceeds 17 °C, its impact tends to reach saturation. Similarly, when rainfall reaches around 20 mm, its adverse effect also approaches the threshold. Temperature is the most important factor affecting the prediction of shared bicycle transfer volume, with temperature, cold weather, and cold forecasts contributing over 35% to the total effect. The interaction effect between temperature and other weather factors accounts for 22% of the total effect.

1. Introduction

With the increasing urban population and the escalating issue of traffic congestion, shared bikes have emerged as a crucial solution to enhance travel efficiency and alleviate traffic pressure [1]. Shared bikes not only reduce the travel volume of private cars and enhance the attractiveness of public transportation but also effectively decrease motor vehicle emissions and energy consumption [2], playing a positive role in improving urban environmental quality and addressing climate change [3]. Additionally, as a flexible and convenient mode of short-distance travel, shared bikes serve as an important choice for transferring to the subway. The combined use of shared bikes and subways has become a significant mode of transportation for commuters. Typically, individuals opt for shared bike rides to reach subway stations in close proximity, followed by subway rides to stations near their destinations. During this journey, shared bikes play the role of the starting point for travel activities. When commuters arrive near their destinations from subway stations, shared bikes serve as the “last mile” solution, providing quick and convenient transfer services [4,5], acting as the endpoint for travel activities. Presently, the combined use of shared bikes and subways is widely adopted in densely populated major cities [6]. acting as the endpoint for travel activities. Presently, the combined use of shared bikes and subways is widely adopted in densely populated major cities [6]. For instance, in the total shared bike travel volume in Shanghai, China, 51% of shared bike travel occurs near subway stations [5,7].

However, shared bike travel is exposed to external environmental conditions and is susceptible to factors such as temperature, rainfall, and air pollution [8,9]. Previous studies have shown that in the Washington, D.C. area, shared bike trips decrease by 15% to 30% on rainy days [10]. Therefore, compared to private cars and conventional public transit, shared bikes are more sensitive to weather changes [11]. Despite the increased flexibility of shared bikes compared to private bikes, for example, allowing on-the-spot parking in rainy weather [8,11], the use of shared bikes still faces certain obstacles. There exists a complex relationship between shared bike travel and weather factors, making shared bike users more vulnerable to the effects of weather changes. Previous studies indicate that [12], compared to driving and using public transportation, shared bike travel is more sensitive to weather variations. Therefore, gaining a deeper understanding of the impact of weather conditions on shared bike travel and the extent of this impact is crucial for grasping the usage patterns of shared bikes under different weather conditions, promoting the combined use of shared bikes and subways, alleviating urban traffic congestion, and reducing carbon emissions.

While existing studies have explored the impact of weather on bike travel [13,14,15], research on the influence of weather on shared bike travel is relatively limited, with even fewer studies focusing on the impact of weather on shared bike transfer volume. In this study, a buffer zone is established centered around subway stations, treating the buffer zone as an area for parking shared bikes when transferring to the subway. The number of shared bike orders ending their journeys within the boundary of this buffer zone is defined as the shared bike transfer volume, serving as the subject of investigation in this study. By integrating data from multiple sources, including subway route information, station geographic data, meteorological data, and air quality index, we conduct a comprehensive study on the impact of weather on the combined use of shared bikes and subways. This extends the scope of previous research on shared bike travel behavior.

It is noteworthy that many studies on the impact of weather on shared bike transfers commonly employ traditional parameter models and linear assumptions. Although these studies can indicate the influence of weather on shared bike transfers, the traditional parameter models and linear models often generate biased estimates and inaccurate interpretations due to the neglect of nonlinear effects and interactions among relevant weather factors [12,16]. For instance, there may be an interaction between humidity and temperature, where changes in humidity can enhance or diminish the impact of temperature. Additionally, the relationship between temperature and shared bike transfer volume may exhibit nonlinear characteristics with certain thresholds [13,17]. To comprehensively explore the nonlinear effects of weather on shared bike transfers, this study uses the K-Means clustering analysis method to construct a research framework that includes basic weather variables, virtual weather variables, and weather forecast variables. Based on this framework, a Random Forest model is employed to investigate the impact of weather on shared bikes trips near subway stations, while interpretable machine learning methods are used to further analyze the effects of weather interactions, revealing the mechanisms through which weather factors influence shared bikes trips, and provide a reference for urban shared bikes trip planning.

2. Literature

2.1. Combined Travel Between Shared Bikes and Subway

Shared bikes, serving as a short-distance travel solution for addressing the “last mile” problem in urban transportation, naturally complement long-distance commuting modes such as the subway. Consequently, numerous studies have explored the integrated use of shared bikes and subway systems. Noland et al. [18] found that shared bike stations near subway stations have higher utilization rates, emphasizing the comprehensive use of shared bikes and subways. Guo et al. [19] using subway daily passenger volume as the independent variable and employing negative binomial regression, identified a positive correlation between subway passenger volume and the volume of shared bike travel. Some literature has constructed a fusion evaluation index for subways and shared bikes, utilizing a generalized additive model to investigate the correlation between this index and land use, social demographics, road design, and transportation facilities [20]. Zhan et al. [21] extracted shared bike data near subway stations and employed local cumulative effect maps to analyze factors related to the integrated use of shared bikes and subways, identifying subway passenger flow and bike density around the service area as the most crucial factors promoting shared bike use near subway stations. Yan et al. [22] extracted data on dockless shared bikes near Shanghai Subway Line 9, visually analyzing the differences in usage patterns of dockless shared bikes around subways on weekdays and weekends, revealing characteristics of dockless shared bike service areas around subways. Data from the Nanjing Subway in China indicated that the coverage area of subway stations determines the demand for the integrated use of shared bikes and subways. Subway stations with higher density reduce the potential for integrated use since shared bikes can be replaced by walking as a connection to the subway [23]. In summary, in recent years, the combined use of shared bikes and subways has shown a rapid growth trend in many regions and has gradually become one of the key directions for the future development of bike-sharing systems. Therefore, it is of great significance to investigate the key factors influencing commuters’ choices to adopt the combined use of shared bikes and subways.

2.2. The Impact of Weather on the Use of Shared Bikes

Previous research has revealed that weather factors significantly impact the use of shared bikes, leading to several scenarios: commuters opting for alternative transportation modes to avoid the effects of weather, delaying their departure until unfavorable weather conditions pass, or choosing not to travel to avoid weather impacts [24]. Therefore, it is crucial to incorporate weather factors into the impact system of shared bike travel. This can be achieved by collecting shared bike road count or order data, along with meteorological data, and conducting modeling and analysis from both temporal and spatial perspectives [25,26], to explore the impact of weather factors on shared bike travel behavior. El-Assi et al. [27] constructed an autoregressive moving average model to estimate shared bike travel volume, finding that warmer weather leads to higher shared bike utilization rates, while rainfall, low temperatures, and humidity decrease utilization rates. Additionally, Wessel studied the influence of predicted weather conditions on cycling volume, revealing that even during rainless periods, predicted rainfall reduces cycling volume by 3.6% [28]. Gebhart and Nolan analyzed shared bike data and discovered that shared bike travelers embarking on journeys farther from subway stations are more susceptible to the impact of rain and low temperatures compared to those starting journeys closer to subway stations [10]. At present, the combined use of shared bikes and subways has become an important component of shared mobility systems. Existing studies on such multimodal travel mainly focus on the influence of built environment factors on the integration of shared bikes and subway systems. For example, Guan et al. evaluated the environmental attributes around subway stations from five dimensions—density, transportation facilities, land use, destination accessibility, and subway ridership—and applied a gradient boosting decision tree model to capture the nonlinear relationship between built environment factors and multimodal travel demand. However, research on the impact of weather conditions on this integrated travel pattern remains relatively limited. This study focuses on examining how weather factors affect the transfer volume between shared bikes and subways, aiming to identify effective analytical methods to explain weather-related influences on transfer behavior. The findings provide valuable insights into the synergistic use of shared bikes and subway systems within multimodal transport frameworks.

2.3. Methods for Predicting Shared Bike Usage

At present, there is a substantial body of research on the prediction of shared bike usage, with the applied methods broadly categorized into statistical approaches and machine learning techniques [29]. Early studies on travel demand prediction primarily relied on time series models, such as the Seasonal Autoregressive Integrated Moving Average (SARIMA) model. SARIMA is a classical statistical time series model well-suited for data with trend and seasonal characteristics, and it is commonly used for daily or weekly demand forecasting. He applied the SARIMA model to forecast short-term demand for public bikes and found that prediction accuracy tended to decline as the forecasting interval increased [30]. To better handle complex spatiotemporal data and uncover nonlinear relationships between variables, researchers have increasingly adopted machine learning models with strong nonlinear fitting and self-learning capabilities for shared bike demand prediction. Long Short-Term Memory (LSTM) is a deep learning model designed for time series analysis, capable of effectively capturing trends, seasonality, and nonlinear patterns in bike usage data [31]. Random Forest (RF), an ensemble learning model, improves prediction accuracy and robustness by constructing multiple decision trees and averaging their outputs. Wang and Cui utilized the RF algorithm to examine the nonlinear effects of built environment factors on bike arrival density across different time periods, demonstrating RF’s ability to manage complex input features and capture nonlinear interactions between variables [32]. In addition, some researchers have introduced interpretable machine learning (IML) methods to enhance model transparency and explanatory power. For instance, Song et al. employed the Gradient Boosting Decision Tree (GBDT) model combined with SHAP to explore the interaction effects of various factors leading to the imbalanced use of DBS at Shenzhen metro stations [33]. Han et al. employed the XGBoost model in combination with the SHAP method to explore key factors influencing shared bike usage [34]. Advanced models based on gradient boosting trees, such as XGBoost, LightGBM, and CatBoost, generally outperform the RF model in terms of prediction accuracy. However, due to their more complex structures, they tend to be less interpretable, which poses challenges for mechanism-based analysis [35]. In light of the above, this study adopts the RF model for its superior interpretability and ability to handle complex feature interactions. Furthermore, it incorporates interpretable machine learning techniques such as Feature Interaction analysis and Accumulated Local Effects (ALE) to systematically investigate the impact of weather conditions on shared bike usage around subway stations.

3. Research Area and Data

3.1. Study Area

The selected study area for this research is Washington, D.C., located in the Mid-Atlantic region on the East Coast of the United States, at the border between the states of Maryland and Virginia. The shared bikes system in Washington, D.C. has been in operation since 2010, making it one of the first cities in the United States to introduce shared bikes. Capital Bikeshare (CaBi) operates the shared bike system in this region, boasting over 8000 shared bikes and more than 800 bike docking stations in Washington, D.C. as of December 2024. Additionally, the region is well-served by an extensive rail transit system, with Washington Subway ranking second nationwide, only behind the New York City subway in terms of passenger volume [36]. As of December 2024, Washington Subway operates six subway lines and 98 stations. Under the planning and development initiatives of the transportation management authorities in the region, the urban shared bike system in Washington, D.C. has gradually emerged as a significant mode for transferring to the city’s rail transit system. The study area, locations of shared bike docking points, and subway routes used in this research are illustrated in Figure 1.

3.2. Data Source and Description

Based on multi-source heterogeneous data from Washington, D.C., spanning from January to December 2022, this study systematically constructs an analytical framework to investigate the impact of weather factors on shared bike transfer volume. The data sources include: (1) Environmental monitoring data: daily PM2.5 concentration statistics (including minimum, maximum, median, etc.) provided by the World Air Quality Index (WAQI), and UV index grading data (ranging from 0 to 11) released by the National Oceanic and Atmospheric Administration (NOAA); (2) Meteorological observation data: weather parameters such as temperature, humidity, wind speed, rainfall, and visibility recorded every three hours by the Ronald Reagan Washington National Airport weather station; (3) Geospatial data: administrative boundaries, rail transit network, and station location data obtained from the OpenStreetMap platform; (4) Operational data: anonymized shared bikes order data provided by the Capital Bikeshare system, including detailed information such as origin and destination coordinates, timestamps of bike usage, and bike type. Detailed characteristics and statistical descriptions of each data source are provided in

Table 1. Dataset source and description.

Dataset	Source	Website	Description
Air pollutant values	World Air Quality Index (WAQI)	https://waqi.info/ (4 May 2025)	Monitoring times, minimum value, maximum value, median, etc.
UV index	National Oceanic and Atmospheric Administration	https://www.nasa.gov/ (6 May 2025)	The index range is 0–11
Historical weather data	Ronald Reagan Washington National Airport weather station	https://www.Rp5.ru (11 April 2025)	Temperature, visibility, humidity, wind speed, rainfall, etc., recorded every three hours
Map information data	OpenStreetMap (OSM)	https://www.openstreetmap.org (5 March 2025)	City administrative region data, POI data, rail transit line network and site data
Shared bike order data	Capital bikeshare (CaBi)	https://capitalbikeshare.com/ (7 April 2025)	User number, site name and its latitude and longitude, starting point and destination latitude and longitude, departure time and end time, user type and bike type, etc.

.

3.3. Study Variables

3.3.1. Extraction of Shared Bikes Order Data near Subway Stations

In order to extract shared bike order data near subway stations, this study utilized ArcGIS (10.8.2) geographic processing software. Washington, D.C. subway station location information was imported, and buffer zones were set to extract shared bike order data that matched the locations within these buffers. However, there is no unified standard for determining the radius of the buffer zone, which may be related to factors such as the distance of subway stations from the city center, subway service capacity, population density, and the composition of surrounding users [26,37]. Referencing previous scholars’ research [3,4,5,7], this study introduced the method of spatial network density to determine the range of buffer zones around subway stations. First, the method assumes that subway stations have a certain degree of attraction to nearby shared bikes, thereby forming a buffer zone. Based on this fundamental assumption, the study needs to establish a function to describe the variation in the density of shared bike orders around subway stations. By progressively expanding the buffer radius and observing the trend in shared bike travel density, an appropriate buffer radius can be determined. Given that Washington, D.C. has a dock-based shared bike system where bikes must be parked at designated stations, for simplification, this study substituted shared bike station density for shared bike order density. The calculation formulas are shown in Equations (1) and (2):

$$f(r) = D_{r,k}/D_{R,k},r = △,2△,...,R$$

(1)

$$D_{r,k} = {\displaystyle \frac{{\displaystyle \sum _{i=1}^{n_k}N(d_{D_i,k} < r + △)}}{\pi {(r + △)}^2}}$$

(2)

In the equations, $f(r)$ represents the spatial density change function of public bike stations near subway station k, with a buffer zone radius of $r$. $△$ is the minimum search step, set to 100 m in this study. $n_k$ is the number of shared bike stations within the maximum buffer zone R, where R is set to 1000 m in this case. $d_{D_i,k}$ is the Euclidean distance from shared bike stations to subway station k, $N()$ is an indicator function. This study takes 41 subway stations in Washington, D.C., as the main research subjects. Figure 2a presents the distribution of values under different buffer zone radii using a boxplot, where the point data represent the spatial density of shared bike stations corresponding to each subway station. It can be observed that when the buffer zone radius expands to 500 m, the mean and variance converge. Therefore, this study sets the buffer distance of subway stations to 500 m and takes the shared bike orders within this 500 m radius as the research subject. Figure 2b displays the kernel density analysis results for shared bike orders near McPherson Square subway station, confirming the 500 m buffer zone as it effectively covers the hotspots of shared bike station distribution.

3.3.2. Dependent Variable

Based on the analysis in Section 3.3.1, this study has successfully extracted all the stations from the Washington Subway rail lines and established a buffer zone with a radius of 500 m. This buffer zone serves as the area where users of shared bikes choose to park their bikes when combining shared bikes with subway travel. In this study, the quantity of shared bike orders that conclude their trips within this buffer zone is defined as the shared bike transfer volume. This variable is set as the dependent variable for the study of combined travel using shared bikes and the subway.

Through the batch processing function of the ArcGIS model builder, 1.04 million shared bike orders in the buffer were extracted, and these shared bike interchanges accounted for one-third of the total shared bike trips. After data cleaning, the spatial and temporal distribution characteristics of the shared bike interchanges are obtained, as shown in Figure 3.

Figure 3a shows the trend of shared bike interchange and total shared bike trips for the whole year of 2022. As can be seen from Figure 3a, the year-round trend of shared bike interchange is more gently distributed than total trips, which indicates that shared bike interchange has formed a stable demand at subway stations and has become one of the main modes of interchange to the subway. Figure 3b shows the trend of shared bike transfer volume in 24 h a day, and there is an obvious phenomenon of morning and evening peaks in shared bike transfer volume. Figure 3c shows the hotspot distribution area of the shared bike transfer volume, and it can be clearly seen that the density of shared bike transfer hotspots in downtown subway stations is much higher than that in suburban areas.

3.3.3. Independent Variables

In order to deeply explore the influence of weather factors on the amount of shared bikes interchange, this study integrates five data sources and K-Means cluster analysis methods to build a three-category weather variable framework to study the role of weather factors on the combination of shared bikes and subway trips. The input variables are mainly three categories: basic weather variables, virtual weather variables, and weather forecast variables, and the statistical descriptions of these input variables are shown in

Table 2. Statistical description of weather independent variables.

	N	Min	Max	Average	SD
Basic weather variables
Visibility	365	7.40	16.00	15.2759	1.62738
Humidity	365	9.10	96.50	62.0710	13.98713
Temperature (Temp)	365	−10.00	30.01	15.0563	9.26532
Wind velocity	365	1.10	14.80	3.9992	1.53985
Temp difference	365	1.60	18.90	9.0345	3.54764
Rainfall	365	0.00	166.90	4.2077	12.53246
UV intensity	365	0.00	10.00	4.4521	2.85377
Air quality	365	0.50	26.40	7.5058	4.86157
Dummy weather variable
Wet weather	365	0.00	1.00	0.1151	0.31954
Dry weather	365	0.00	1.00	0.1676	0.37401
Cold weather	365	0.00	1.00	0.1319	0.33881
Wind weather	365	0.00	1.00	0.1126	0.31658
Rainy weather	365	0.00	1.00	0.1374	0.34470
Strong UV weather	365	0.00	1.00	0.1154	0.31993
Weather forecast variables
Forecast of rain	365	0.00	1.00	0.2932	0.45583
Forecast of UV warning	365	0.00	1.00	0.2060	0.40502
Forecast of low temp	365	0.00	1.00	0.0549	0.22819

.

(1)

Basic weather variables

In previous studies, numerous scholars have found that environmental factors such as temperature and rainfall have a significant impact on bike travel [19,38,39]. Therefore, in this study, temperature, humidity, visibility, wind speed, rainfall, and values of UV index and air pollutants from January 2022 to December 2022 were included in the study, and they were considered as the underlying weather variables. Since some of the original environmental statistics are minute-by-minute, hour-by-hour, and day-by-day data with inconsistent data statistical intervals, the base weather variables were homogenized in this study.

(2)

Dummy weather variable

In order to investigate the influence of unfavorable weather on shared bikes interchange, this study constructs a virtual weather variable based on the basic weather variable, which contains six kinds of dry, humid, low temperature, strong wind, rainfall and strong ultraviolet rays. The virtual weather variables are dichotomous variables with values of 0 and 1. Due to the selection method of the virtual weather variables, there are mainly two kinds of definitions, absolute value definition and relative value definition, absolute value definition refers to the method of taking a fixed value, e.g., defining the temperature lower than 0 °C in a day as low temperature weather; relative value definition is to take a percentile method, e.g., arranging the temperature values according to the size and taking the value in the lowest 20% of the values. portion is defined as cold weather [40,41,42,43]. These two selection methods of virtual weather variables have their own advantages. However, they still cannot accurately reflect the specific nodes of the changes in the characteristics of the virtual weather variable observations. In this study, the basic weather variable data were divided into five categories, and the K-Means clustering algorithm shown in Equations (3) and (4) was introduced [44]. to solve the clustering centers by minimizing the variance within the same cluster, maximizing the cluster distances of different clusters, and solving the cluster center values, and then, the highest or lowest cluster center value is selected as the boundary center value for dividing the virtual weather variables.

$$E = {\displaystyle \sum _{i=1}^k\sum _{x\in C_i}\left|\right|x - {\mu }_i|{|}^2}$$

(3)

$${\mu }_i = {\displaystyle \frac{1}{\left|C_i\right|}}{\displaystyle \sum _{x=C_i}x}$$

(4)

where k denotes the number of clusters, ${\mu }_i$ is the average value of observation x in class $C_i$, also called the cluster center value. E is the sum of error squares, the smaller the value, the closer the observation x is to the cluster center, resulting in a better clustering effect. After the iterative calculation of the K-Means clustering algorithm, the clustering results of virtual weather variables shown in

Table 3. K-Means clustering results.

Type	I	II	III	IV	V
Humidity
Cluster center value	0–41%	41–54%	54–65%	65–74%	74–85%
sample size	63	103	89	68	42
Temperature
Cluster center value	0–6	6–12	12–18	18–25	25+
sample size	48	55	68	66	116
Wind velocity
Cluster center value	0–1.7	1.7–2.6	2.6–3.5	3.5–4.8	4.8–6.9
sample size	23	63	114	87	41
Rainfall
Cluster center value	0	0–2	2–16
sample size	242	72	51
UV intensity
Cluster center value	1	2	3.5	7	9.1
sample size	37	69	83	92	42

are obtained. Among them, the observed values of rainfall are clustered into 3 major categories, and the rest of the virtual weather variables are clustered into 5 major categories. Based on the K-Means clustering results shown in Table 3, this study defines the Type I and Type V samples of humidity as dry weather and wet weather, respectively; the Type I samples of temperature as cold weather; the Type V samples of wind speed as wind weather; the Type III samples of rainfall as rainy weather; and the Type V samples of UV intensity as strong UV weather.

(3)

Weather forecast variables

Most of the previous studies only consider the effects of historical weather or real-time weather on bike sharing ridesharing, but in real life, travelers often also refer to the weather forecast when planning their trips and deciding on their future travel modes [36,45]. Since unfavorable weather events such as low temperatures, strong UV rays and rainfall can have a prospective impact on shared bikes interchange, it is necessary to include low-temperature forecasts, UV warnings and rainfall forecasts as weather forecast variables in this study.

4. Research Methods

4.1. Random Forest Model

Random Forest (RF) algorithm is a classical ensemble learning method and an extension of Bagging [46]. It further introduces random attribute selection into the training process of decision trees, building on the foundation of constructing a Bagging ensemble with decision trees. Compared to traditional regression models, RF has several advantages. Firstly, RF exhibits a certain degree of robustness as it imposes no restrictions on the type of input variables; they can be numerical, categorical, continuous, or discrete [47]. Moreover, RF is insensitive to skewed distributions, outliers, and noisy data. Secondly, RF requires relatively fewer hyperparameter tuning. The main hyperparameters include the number of decision trees, the maximum number of features to consider when constructing a decision tree, and the maximum depth of the decision tree. These hyperparameters are not very sensitive to their values. Importantly, RF, as a tree-based ensemble learning algorithm, possesses strong nonlinear fitting capabilities due to its flexible modeling structure. It can automatically learn the correlation and weight distribution among input features, model complex nonlinear relationships between input variables and response variables, and capture high-order interactions between variables.

In recent years, some progress has been made with the Interpretable Machine Learning (IML) [48,49], the combination of IML and Random Forest (IML+RF) has found widespread applications in the field of transportation, serving various purposes such as predictions and data interpretation. For instance, RF has been employed for short-term forecasting of bus passenger volumes, and IML has been utilized to assess the influencing factors on bus travel speeds [50,51]. Additionally, RF has been used to establish regression models between pollutant concentrations, while IML has been applied to analyze the relative importance of variables such as meteorological conditions, time factors, and traffic flow [52]. This suggests that employing IML+RF for the analysis of shared bike transfer volumes is feasible. Therefore, this study first utilizes RF to establish a regression model for shared bike transfer volumes based on comprehensive weather variables. Subsequently, employing IML algorithms for feature interactions, feature importance ranking, and cumulative local effect plots, the study interprets the results of the regression model.

4.2. Model Interpretability

4.2.1. Feature Interaction

There may be interactions between weather factors, and when exploring the effect of weather on shared bikes to the subway, it is not possible to consider only the sum of the effects of each part of the weather variables. This is due to the fact that the effect of the influence of one weather variable is often also dependent on the value of another weather variable. Therefore, this study will explore the interaction effects between weather variables using Feature Interaction from the IML package. Feature Interaction was proposed by Friedman and Popescu [53] and is based on the H-statistic. The H-statistic value is based on the underlying theory of decomposition of the partial correlation function, which is defined as a function that depends only on the characteristics of $x_j$ and $x_k$ as well as the interaction between the $x_j$ and $x_k$ characteristics, and when the two characteristics do not interact with each other, the partial correlation function can be decomposed into the form of Equation (5):

$$P{D}_{jk}(x_j,x_k) = P{D}_j(x_j) + P{D}_k(x_k)$$

(5)

When there is an interaction between two features, the H-statistic value for assessing the strength of the interaction is calculated by Equation (6):

$$H_{jk}^2 = {\displaystyle \frac{{\displaystyle \sum _{i=1}^n{[P{D}_{jk}^2(x_j^{(i)},x_k^{(i)}) - P{D}_j(x_j^{(i)}) - P{D}_k(x_k^{(i)})]}^2}}{{\displaystyle \sum _{i=1}^{n}P{D}_{jk}^2(x_j^{(i)},x_k^{(i)})}}}$$

(6)

where $P{D}_{jk}(x_j,x_k)$ is the bi-directional partial correlation function of the two features J and K, $P{D}_j(x_j)$ and $P{D}_k(x_k)$ are the partial correlation functions of the individual features. i is the index of the sample, $P{D}_{jk}({x_j}^{(i)},{x_k}^{(i)})$ is the bi-directional partial correlation function of features J and K for the i-th sample, and $P{D}_k({x_k}^{(i)})$ is the partial correlation function of feature K for the i-th sample.

4.2.2. Importance of Features

To elucidate the impact of weather factors on shared bike transfer volumes, this study introduces the parameter of feature importance, utilizing it to assess the significance of weather factors. Feature importance is gauged by calculating the variation in model prediction errors based on changes in feature values. If altering a particular feature value leads to an increase in the model’s prediction error, the feature is deemed “important” [54,55]. Conversely, if the error decreases, the feature is considered “unimportant.” In contrast to linear regression models, which only reflect the significance levels of variables, feature importance comprehensively considers both main effects and interaction effects of feature variables. It ranks the importance of feature variables, providing policymakers with deeper insights when seeking intervention measures.

4.2.3. Accumulated Local Effects Plot

The Accumulated Local Effects plot (ALE) is commonly used to illustrate how explanatory variables influence the predictions of a machine learning model. In comparison to Partial Dependence Plots (PDP), ALE plots serve as a faster and unbiased alternative. PDP assumes that all explanatory variables are independent, making it unsuitable when strong correlations exist among explanatory variables [56]. In this study, since weather factors are the explanatory variables and there is substantial correlation among most weather factors, PDP is not suitable. Therefore, ALE plots are considered a feasible explanatory tool for weather variables in this research.

4.3. Model Evaluation

Typically, metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R²) are commonly used to evaluate model performance. While both RMSE and MAE can be employed to measure model error, RMSE is more sensitive to larger prediction errors compared to MAE, as RMSE’s errors are squared before averaging. The R² value ranges from 0 to 1, reflecting the variation in the response variable explained by the model. When a model exhibits the lowest RMSE and MAE, as well as the highest R², it indicates that the model minimizes the difference between predicted values and actual values, achieving optimal training results. The formulas for calculating R², MSE, and RMAE are provided in Equations (7)–(9):

$$R^2 = {\displaystyle \frac{{\displaystyle \sum _{i=1}^n{(y - \overline{y})}^2} - {\displaystyle \sum _{i=1}^n{(y - y_{pre})}^2}}{{\displaystyle \sum _{i=1}^n{(y - \overline{y})}^2}}}$$

(7)

$$MSE = {\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n{(y - y_{pre})}^2}$$

(8)

$$RMSE = \sqrt{MSE}$$

(9)

where $y_{pre}$ represents the predicted count of shared bike transfers, $y$ represents the actual count of shared bike transfers, and $\overline{y}$ denotes the mean count of actual shared bike transfers.

5. Results Analysis

5.1. Comparative Analysis of Results Based on Different Models

To evaluate the performance of the Random Forest model, this study integrated shared bikes order data with weather data into a combined dataset, which was then split into 70% for training and 30% for testing. Subsequently, we determined the setting of 1000 decision trees using cross-validation, set the random seed to 50 to ensure consistent output, performed model parameter tuning using grid search, and trained the RF model on the training set. Subsequently, we determined the setting of 1000 decision trees using cross-validation, set the random seed to 50 to ensure consistent output, performed model parameter tuning using grid search, and trained the RF model on the training set. Figure 4 illustrates the performance of the Random Forest model. By comparing the predicted values with the observed values for both the training and testing sets, it can be observed that the predicted shared bikes transfer volume is relatively close to the observed values in the original dataset. The obtained R² and RMSE for the training set are 0.81 and 525.50, respectively, indicating that the Random Forest model can effectively predict shared bikes transfer volume under the influence of weather conditions.

5.2. Interaction Effects of Weather Factors

Figure 5 displays the global interaction effect strength rankings for all weather variables and a heatmap illustrating the second-order interaction effect strengths between weather variables. In terms of global interaction effect strength values, except for temperature, the interaction effect strengths of other weather variables are relatively weak, with H-statistic values generally below 10%. The H-statistic can be understood as the proportion of weather variable interaction effects in the overall effect. The interaction effect of temperature is more pronounced than that of other weather variables, indicating that the impact of temperature on shared bikes transfer volume is relatively unstable. Changes in the values of other weather variables may alter the magnitude and direction of the temperature’s impact on shared bikes transfer volume.

Categorically, the interaction effects of basic weather variables are most evident, while virtual weather variables and weather forecast variables exhibit weaker interaction effects. This is because virtual weather and weather forecast variables have strong main effects, and the impact of these two variable types on shared bikes transfer volume is consistently negative and relatively stable. Regarding the second-order interaction effects of weather variables, there is a significant interaction between visibility and air quality, as well as between visibility and rainfall. This phenomenon aligns with real-life experiences, where heavy rainfall and poor air quality can both reduce visibility. Thus, when low visibility coincides with heavy rainfall or high air pollution, the overall effect cannot be represented as the additive impact of multiple weather factors due to the existence of interaction effects. In addition, there are also significant second-order interaction effects between related weather variables such as temperature and humidity, rainfall and wind, temperature difference and visibility, as well as temperature difference and air quality.

5.3. Ranking the Importance of Weather Factors

Figure 6 illustrates the Relative Importance (RI) rankings of all weather variables. This figure provides a clear reflection of the impact of each weather variable on the prediction of shared bikes transfer volume. Higher RI values indicate that the corresponding weather factors are more crucial for accurately predicting shared bikes transfer volume. The results indicate that the most important basic weather variables affecting shared bikes transfer volume are temperature, UV intensity, humidity, rainfall, visibility, wind speed, temperature difference, and air quality. Temperature-related variables play a crucial role in shared bikes transfer volume, with temperature, low-temperature weather, and low-temperature forecasts contributing to 35% of the prediction. Clearly, rising temperatures in summer attract more people to use shared bikes for subway transfers in comfortable weather, while in winter, the usage of transportation modes directly exposed to the external environment significantly decreases. Humidity, visibility, wind speed, and air quality also have a significant impact on shared bikes transfer volume because they determine the safety and comfort of the shared bikes’ riding environment.

Investigating the impact of adverse weather from the virtual weather category, contrary to previous perceptions, low-temperature weather (RI = 9.7) has a much greater impact on predictions than rainy weather (RI = 3.7), indicating that the adverse effects of low temperatures on shared bikes transfer volume are much greater than those of rainy weather. The adverse impact of rainy weather on shared bikes transfer volume is close to that of strong wind weather (RI = 3.1), while the impact of strong UV weather is the weakest (RI = 1.6).

In the weather forecast category, the importance ranking shows that rainfall forecasts (RI = 1.1) have the lowest contribution rate to predictions. This may be due, in part, to the lower accuracy of rainfall forecasts compared to low-temperature forecasts and UV warnings, and, on the other hand, the overall weaker impact of non-real-time weather on shared bikes transfer volume.

5.4. Impact of Weather Factors on Shared Bikes Transfer Volume

This study utilizes Accumulated Local Effects (ALE) plots to interpret the complex nonlinear relationships between the predicted values of shared bikes transfer volume and weather variables, as well as the interactions among weather variables. For simplification purposes, this research focuses on analyzing the main effects of some fundamental weather variables and their additional interaction effects.

(1)

Temperature and Humidity

As shown in Figure 7a,b, temperature exhibits a positive correlation, while humidity shows a negative correlation with the shared bikes transfer volume. An increase in temperature significantly promotes the combined use of shared bikes and subway transportation. As the temperature rises from 0 °C to 17 °C, the shared bike transfer volume shows an almost linear increasing trend with temperature, with the predicted transfer volume increasing by more than 1500 trips, representing a growth of over 75%. However, when the temperature exceeds 17 °C, its impact effect approaches saturation. In contrast, the comprehensive impact of humidity is opposite to temperature. When humidity increases from 10% to 60%, the change in shared bike transfer volume is relatively gradual. Beyond 60%, the shared bikes transfer volume sharply decreases.

Additionally, there is a strong interaction effect between temperature and humidity. According to the second-order interaction effect ALE plot of humidity and temperature on the predicted shared bikes transfer volume shown in Figure 7c, the darker shaded ALE values above 0 indicate additional positive effects, while lighter shaded ALE values below 0 indicate additional negative effects. This reveals the interaction between temperature and humidity: in cold and humid weather, the predicted shared bikes transfer volume is subject to additional positive effects. This indicates that under cold and humid weather conditions, the predicted shared bikes transfer volume is influenced by an additional positive interaction effect. This may be due to the complex impact of temperature and humidity on shared bikes transfer volume, where the combination with subway transfers could actually enhance travel demand. This reveals the flexible coping strategies people adopt in multimodal transport systems when facing adverse weather, especially when the need for travel efficiency persists, making short-distance shared bikes still an effective tool for connecting with subway transfers.

(2)

Rainfall and Wind velocity

In addition to temperature and humidity, both rainfall and wind speed also have significant impacts on the shared bikes transfer volume, jointly influencing the safety and convenience of the riding environment. Figure 8a,b illustrate the relationships between rainfall and wind speed with the shared bikes transfer volume, showing negative correlations in both cases. The adverse impact of increasing rainfall on shared bikes transferring to the subway is pronounced, with an approximate reduction of 20 transfers for every 1 mm increase in rainfall. When the rainfall reaches around 20 mm (meteorologically defined as moderate rain), its adverse effects approach a threshold. In comparison to rainfall, the influence of wind speed on shared bikes transfer volume exhibits a stepped decline. In the stage where wind speed is less than 5 m/s (meteorologically defined as a breeze), there is a weak facilitating effect on shared bikes transfer to the subway.

Figure 8c,d, respectively, depict the second-order interaction effects between rainfall and wind speed with temperature. It is noteworthy that at a temperature of −10 °C, regardless of the numerical value of rainfall, there is an additional facilitating effect on the prediction of shared bikes transfer volume. This phenomenon is due to the fact that rainfall statistics also include snowfall when the temperature is below 0 °C, essentially making rainfall equivalent to snowfall in cold weather. Therefore, snowfall in cold weather mitigates the adverse impact of low temperatures on shared bikes transfer volume. The interaction effect between wind speed and temperature is divided into upper and lower regions based on wind speed values. When the wind speed is below 7 m/s (meteorologically defined as a light breeze), an increase in temperature is associated with an increase in the predicted shared bikes transfer volume. However, when the wind speed exceeds 10 m/s, the real data in this region is sparse, rendering ALE estimates in these areas less reliable.

(3)

Temperature differences, visibility and air quality

Temperature difference, visibility, and air quality also influence the integrated use of shared bikes and subways. As shown in Figure 9a, there is a nonlinear effect between temperature difference and the predicted shared bikes transfer volume. When the temperature difference is less than 10 °C, it exhibits a positive correlation with the shared bikes transfer volume; however, when the temperature difference exceeds 10 °C, it shows a negative correlation with the shared bikes transfer volume. When the temperature difference in the Washington area is 10 °C, shared bike usage is at its highest. When the temperature difference in the Washington area is 10 °C, shared bike usage is at its highest. When the temperature difference in the Washington area is 10 °C, shared bike usage is at its highest. Figure 9b illustrates a positive correlation between visibility and the predicted shared bikes transfer volume, indicating a close-to-linear effect between the two. Regarding air quality, as the air pollution index increases, the predicted shared bikes transfer volume shows a linear decreasing trend.

Figure 10b and Figure 10c, respectively, present the second-order interaction effects between air quality and (b) temperature and (c) humidity. When the temperature is in the range of 0–10 °C and a high pollution index is present, the predicted shared bikes transfer volume experiences an additional negative impact. On the other hand, when air quality is good and the environment is humid, the predicted shared bikes transfer volume receives an additional positive impact.

Table 4. Impact of weather conditions on bike shares: A comparison between previous studies and this study.

Authors	The Impact of Weather on Bike-Sharing Usage.	The Impact of Weather on Shared Bikes Transfer Volume
Xie et al. (2022) [57]	Temperature, thunderstorms, and wind speed significantly affect bike-sharing usage on Saturdays and Sundays.	An increase in temperature significantly promotes the combined use of shared bikes and subway transportation.
Liu. (2022) [58]	Humidity has a relatively small impact on bike-sharing usage within a certain range.	When humidity increases from 10% to 60%, the change in shared bike transfer volume is relatively gradual. Beyond 60%, the shared bike transfer volume sharply decreases.
Liu & Fan (2024) [59]	Temperature difference exhibits a nonlinear effect on bike-sharing usage.	When the temperature difference is less than 10 °C, it exhibits a positive correlation with the shared bikes transfer volume; however, when the temperature difference exceeds 10 °C, it shows a negative correlation with the shared bikes transfer volume.
Bai et al. (2024) [60]	Long-term cycling of shared bikes under poor air quality may pose health risks, leading to a decrease in bike-sharing usage.	as the air pollution index increases, the predicted shared bikes transfer volume shows a linear decreasing trend.
Zhao (2021) [61]	Weather temperature, rainfall, wind speed, and air quality have a significant impact on the demand for bike-sharing.	the relationships between rainfall and wind speed with the shared bikes transfer volume, showing negative correlations in both cases.
Xie (2023) [62]	Low temperatures, poor visibility, and excessive rainfall all lead to a decrease in overall bike-sharing usage.	Visibility shows a positive correlation with the predicted shared bikes transfer volume

presents the impact of weather on bike share usage as reported in existing studies, as well as the impact identified in this study. Although there are differences in study regions and time periods, the significant influence of weather on bike share usage is still evident, highlighting the distinctions between weather effects on general bike shares and on the combined travel patterns explored in this study.

6. Conclusions

This study comprehensively analyzed the impact of weather factors on the transfer volume of shared bicycles near a subway station by studying the order data and meteorological data of shared bicycles near a certain city’s subway station; By using random forest models and interpretive machine learning methods, the important determining factors of shared bicycle transfer volume were ranked, revealing various nonlinear, threshold effects, and interaction effects among them. The main conclusions are as follows:

(1)

The random forest model has shown high accuracy in predicting the impact of weather changes on shared bicycle transfers. Machine learning algorithms can clearly explain the relationship between weather factors and shared bicycle transfers. It is feasible to use the RF+IML method to study the impact of weather variables on shared bicycle transfers.

(2)

The interaction effect mainly occurs between basic weather variables, especially between temperature, humidity, rainfall, and wind speed. Temperature is the most important factor affecting the prediction of shared bicycle transfer volume, with temperature, low-temperature weather, and low-temperature forecasts contributing over 35% of the total effect. The interaction effect between temperature and other weather factors accounts for 22% of the overall effect.

(3)

The relationship between temperature, humidity, and rainfall and the number of shared bicycle transfers often has specific activation and threshold effects. When the humidity is less than 60%, the change in the transfer volume of shared bicycles is relatively gentle. After exceeding 60%, the transfer volume of shared bicycles sharply decreases; Once the temperature exceeds 17 °C, its effect approaches saturation; When the rainfall reaches about 20 mm, its adverse effects approach the threshold.

(4)

The interaction effects between weather factors have a significant impact on shared bike transfer volume. Under cold and humid weather conditions, shared bike transfers are positively influenced by interaction effects. When the temperature in the Washington area approaches −10 °C and rainfall occurs, the predicted transfer volume of shared bikes generally shows an additional promoting effect. When wind speed is below 7 m/s, an increase in temperature is significantly associated with an increase in predicted transfer volume. In contrast, when the temperature is in the range of 0–10 °C and the air pollution index is high, the predicted transfer volume experiences an additional negative impact. On the other hand, under conditions of good air quality and high humidity, the predicted shared bike transfer volume receives an additional positive gain.

To mitigate the adverse effects of unfavorable weather on shared bike-to-subway transfer behavior, this study proposes the following policy recommendations: (1) Optimize infrastructure: Install covered bike-sharing parking areas around subway stations to reduce the impact of rain, snow, and other severe weather conditions on cycling behavior, thereby improving travel convenience and comfort; (2) Establish weather alert and travel guidance mechanisms: Utilize real-time meteorological data to provide users with advance cycling recommendations, dynamically guiding them to select the most optimal transfer modes based on weather changes. Integration with mobility apps can offer alternative routes or transportation suggestions; (3) Enhance intelligent dispatching systems: Based on predicted shared bike demand, optimize the operational dispatching strategy in advance. Before the onset of adverse weather, adjust bike distribution and capacity allocation to ensure the stability of travel services. This study investigates the impact of weather conditions and their interaction effects on shared bike usage near subway stations in the Washington, D.C. area. Future research could be extended to multiple cities and diverse climate zones to further explore how different climatic environments influence shared bike usage behavior around subway stations.