Introduction

The development of knowledge and technology has greatly influenced the field of cartography, leading to more effective and intuitive use of maps. However, difficulties and interpretation errors may still arise in the process of map reading (Medyńska-Gulij 2010). Understanding factors that contribute to successful map reading can provide map designers and cartographers with valuable knowledge. This may include studying the design of the map and its legend (Brychtová and Çöltekin 2014; Edler et al. 2020) as well as the user’s visual strategy and perception (Gilmartin 1981; Gong et al. 2022; Philips and Noyes 1982). By identifying the most effective and efficient ways to read maps, we can improve the usability and accessibility of cartographic products for the wider range of users.

Furthermore, with the increasing availability of digital and mobile cartographic products, it is becoming even more important to understand how users interact with these products and how they can effectively and efficiently find the information they need. Currently, almost every mobile map has basic interaction capabilities, such as zoom or pan. The interfaces take the form of a graphical user interface (GUI). In terms of mobile devices, their interface is fundamentally different from the one in desktop devices. Among others, the work of Muehlenhaus (2014) devoted a lot of attention to this issue.

Some interactive activities, such as zoom or pan, do not have typical graphical representations on mobile maps as they do on their desktop equivalents. Some interactions are triggered intuitively by pinch or double tap, which is considered as a part of the adaptation to a mobile device (Reichenbacher 2001; Nivala and Sarjakoski 2007; Medyńska-Gulij et al. 2022). Researchers take into consideration not only map design but also the interface and interactivity design, depending on the map usage context, which assumes the reconfiguration of, for instance, GUI parameters, such as temporal navigation bar or button localization, according to user preferences (Horbiński et al. 2020; Bartling et al. 2022).

However, not all interactions are deprived of graphical representations. Temporal maps add the additional dimension to cartographic products, which can be represented by date or time sliders (MacEachren 1994; Cybulski 2016). These sliders allow users to navigate through the timespan of a cartographic animation on a mobile device. While some interactions with spatio-temporal visualizations may not have any graphical representations, the use of sliders and buttons to navigate through time is a well-established and useful way to interact with temporal maps (Roth 2013). In fact, the use of interactive timelines and other similar interactive elements can enhance the user experience and help users understand complex temporal data better.

So far, the evaluation of effectiveness and usability of mobile GUI and user experience (UX) has been based on web maps, without considering additional temporal dimension. Such evaluation is often grounded in a task-based performance (Nivala et al. 2008; Bartling et al. 2021a). Therefore, usability is related to effectiveness (or success rate) and efficiency, and, on other hand, to the comfort of use. In the context of mobile cartography, user experience (UX) refers to the overall quality of interaction that users have with a mobile mapping application or service. It encompasses the users’ perceptions, feelings, and satisfaction with the mobile cartography experience (Bartling et al. 2021).

On the other hand, users’ visual experience refers to how users perceive and interact with visual elements such as interactive buttons. It is a critical component of overall UX, as it can greatly impact user understanding of the information presented on the map (Coleman 2017). More frequently, recording users’ eye movement (or, at least, carrying out future studies in this field) and analyzing visual strategy with the use of eye trackers is required (Horbiński et al. 2021; Ruginski et al. 2022). This type of analysis is based on fixations, i.e., relatively stable gaze locations, and saccades, i.e., rapid movement of eye balls (Fisher and Ramsperger 1984; Krassanakis and Cybulski 2021). The eye-tracking analysis could help cartographers and map designers to understand user experience and facilitate multidimensional, interactive, and dynamic maps.

In the context of cartography and map design, “user experience” refers to how users perceive and interact with maps or mapping applications. It encompasses factors such as ease of use, efficiency in achieving tasks, visual appeal, and overall satisfaction. Understanding user experience is crucial for map designers because it helps them create maps that are not only functional but also enjoyable and efficient for users. “Multidimensional, interactive, and dynamic maps” refers to maps that go beyond traditional static representations. The study aims to understand how users experience and interact with maps that are not static but instead incorporate multiple layers of information, allow user interaction, and provide real-time or dynamic data updates. The study’s findings can then be used to improve the design of such maps, enhancing user experience and usability.

Weather maps are commonly used for visualizations and communicating weather information to users. However, the increasing use of mobile phones to get access to weather information has resulted in the need for weather maps that are optimized for mobile phone screens. In particular, the navigation of time-based weather data on mobile phones presents a challenge for map designers, as the limited screen space can make it difficult for users to perceive changes in weather patterns over time. The main aim of this paper is to assess the effectiveness and efficiency of different types of GUI during map-based tasks, and users’ visual experience of interactive temporal navigation on mobile map. To accomplish this task, we designed three versions of GUI, which enable temporal navigation and map-based tasks on a mobile device. To help understand the UX issues related to specific user’s behavior during interactions with the map, we included eye movement recording technology and interactions analysis. In relation to the mentioned issues, some research questions arise:

  1. 1.

    How do different GUI designs impact users’ visual strategies and task performance when navigating time-based weather data on mobile maps?

  2. 2.

    What are the effectiveness, efficiency, and user experience implications of various temporal navigation interfaces (GUI designs) for map-based tasks on mobile devices?

  3. 3.

    What insights can eye-tracking technology provide regarding user behavior and visual attention during map-based tasks on mobile weather maps?

Addressing these research questions will help map designers and cartographers understand the user’s visual experience and behavior while using mobile weather maps better. The knowledge can be used to optimize the design of GUI layouts for temporal navigation, improving the usability and effectiveness of weather maps on mobile devices. Additionally, by using eye movement recording technology and interaction analysis, designers can gain insights into how users interact with different GUI designs and identify areas for improvement in mobile weather map design.

Related Work

GUI Design Impacts User Experience and Performance

Assessment of mobile maps involves the usability of the design. Since it is about not only graphical design of map components, such as roads and rivers, but also the adaption of interactivity to a smaller device. A mobile map that is useful is not only properly designed graphically. It should be easy to handle in terms of interactions. In this context Bartling et al. (2021b) showed that properly designed interaction buttons helped users in self-localization map-based tasks and made them confident in terms of map use. They also proposed objective measures such as success rate and subjective measures such as user comfort and confidence to evaluate the usability of non-temporal interactive maps.

Another study assessing maps on mobile devices was presented by Dillemuth (2005), which suggested that user preferences and different contexts of usage should be taken into account when investigating effective mobile map design. The study involved a pedestrian route-following task, and interactions with the mobile device were measured. The mobile-first approach, as advocated by Ricker and Roth (2018), emphasizes the importance of initially designing maps for smaller screens and subsequently adapting them for larger screens. This approach is underpinned by the recognition that mobile devices possess smaller screens and distinct interaction capabilities compared to desktop devices, potentially influencing users’ interactions with the map.

Another research study based on assessment of mobile interactions was conducted by Horbiński et al. (2020). In that work, researchers asked participants about the location of specific buttons, such as geolocation, zoom, search, or route find. Their motivation resulted from studies focused on users’ preferences, known as egocentric design (Meng 2005). Results supported the thesis that interactions on the mobile phone should be placed differently from desktop solutions since mobile devices are handled differently (e.g., the thumb is used during interactive activities). Those studies were followed by the research that involved empirical studies with users (Horbiński et al. 2021). They compared users’ preferred GUI arrangement in map-based task on mobile and desktop devices. The study showed that participants’ habits of using windows-based interactions (the crucial interactive buttons in the upper bar) were also transferred to mobile interactions. Therefore, in designing new cartographic products for mobile display, one should consider not only pure users’ preferences but also their habits.

With the increasing use of mobile devices in our perception of places, Roth et al. (2018) presented a study that aims to provide insights for creating and utilizing a mobile map for teaching and learning. Some findings suggest that key design considerations are related to user-map interactions. The studies mentioned earlier provide compelling evidence that designing an effective mobile map requires an intuitive interface that strengthens the cognitive connection between the map and the depicted phenomenon.

However, these studies failed to take the temporal dimension into account which is crucial for weather forecast maps.

Implications of Temporal Navigation Through Spatial Information

Reichenbacher (2001) addresses the issue of time in mobile cartography, which necessitates additional interactions and data representation. Therefore, Harrower and Fabrikant (2008) proposed the three most common kinds of temporal legends that are now widespread in the cartographical design of animated and spatio-temporal interactive maps. These are a digital clock, which shows a specific moment in time; cyclical, which allows user to select a specific moment on a timeframe; and a bar, which allows one to move rapidly between the timespan of animation. Selected ways of time navigation were originally designed for desktop displays. Some of these interactive solutions were part of the research agenda during desktop studies. Different levels of interactivity for specific tasks helped reduce cognitive load and address some perceptual problems (Midtbø et al. 2007).

However, increasing importance of mobile devices suggests that these methods of interaction should be empirically verified in order to remain useful and efficient. For example, the digital clock legend may be too small to be read on a mobile device while the cyclical legend may require too much interaction and screen space. It was also discussed in the study of Andrienko et al. (2003). On the other hand, the bar legend may be more suitable for mobile devices, as it allows users to quickly navigate through the timespan of the animation without the need for too much interaction or screen space.

Therefore, the design of a suitable interactive legend for mobile devices has not been thoroughly investigated, despite the rapid growth in the popularity of mobile maps. Mobile multimedia features like animations or virtual flybys enable users to swiftly identify objects and possess visual appeal. However, without adequate time control, they can potentially overwhelm users cognitively (Meng 2008). Especially when weather forecast maps involve integrating spatial data with temporal aspects, which can be challenging for certain map-based tasks. Participants tend to have better comprehension of spatial data rather than temporal data (Cybulski and Medyńska-Gulij 2018).

Weather forecasts are currently available through mobile apps, and some of them offer map-centered or multimodal designs (Zabini 2016). However, Abraham et al. (2015) research showed that users prefer verbal and numeric information over cartographic representations. Despite this preference, the number of weather forecasts using various types of maps is still growing (Nagle 2014). On the other hand, the use of graphic-text pop-up information in the weather map interface led to an improvement in participants’ performance in Shao et al. (2022) study. Therefore, there is a need to study weather map design and interactions to gain a better understanding of users’ decision-making and comprehension of the spatiotemporal dimensions behind meteorological data.

Insights from eye-tracking technology

Weather maps were evaluated with the use of eye-tracking technology in the study conducted by Popelka et al. (2019). The aim of the study was to analyze how web weather maps were used and perceived in terms of design aspects. In the research, several maps were evaluated. Most of selected web-based weather maps enabled time navigation only through a time slider and/or buttons back and forward. Both amateur and expert users participated in this study. However, their interests were different since experts were more explorative than amateur users, who focused mainly on the map content. It also showed that some maps that seemed attractive in the first glance caused some difficulties in exploration. Researchers also pointed out the problem of interacting with map through controls. Users needed some time and practice in order to find time navigation easy. That rose an issue of user-map interactions.

This becomes even more crucial when developing and evaluating a weather map app for visually impaired users (Weir et al. 2012). Therefore, as Craig (2015) suggests, the proper implementation of interaction mechanisms can overcome many limitations associated with mobile devices, potentially serving as a key factor in the effective visualization of temporal data. Other conclusions by Hegarty et al. (2010) suggest that in relatively simple map-based tasks, there could be a high influence of display factors on participants. On the other hand, fixations in weather map perception are directed by top-down factors such as additional knowledge. That could be one of the factors that was crucial for understanding users’ preferences on mobile map display.

In summarizing existing research studies, it becomes evident that assessing the usability of mobile maps, with a focus on interactive design, is of paramount importance. Various studies have underscored the significance of factors such as user preferences, weather map design considerations, and adopting a mobile-first design approach. However, a research gap is discerned in terms of addressing the temporal dimension, especially in the context of designing interactions for weather forecast maps. Additionally, there is a pressing need for empirical validation of time navigation methods tailored for weather forecast maps on mobile devices.

Methodology

Materials

We designed 3 interactive temporal navigation interfaces for exploration of spatio-temporal forecast weather data. They are related to the types suggested by Harrower (2003) and Harrower and Fabrikant (2008). Firstly, slidebar navigation (named in this study), consisting of an interactive bar, which users may touch and move to the selected time period. The slidebar navigation interface is characterized by the presence of an interactive bar that users can manipulate to navigate through different time periods. This interactive bar typically represents a timeline, allowing users to slide it back and forth to select a specific moment in time. The key feature of this temporal navigation interface is that users can interact with the slidebar by touching and moving it horizontally on the mobile device’s screen. The slidebar allows users to precisely select a particular point in time within the available timeframe. As they slide the bar, they can see the associated date and time, which provides them with temporal context.

Secondly, button navigation, which consists of two buttons that enable users to move back and forth through different time periods. Users can interact with this interface by tapping the navigation buttons on the mobile device’s screen. In the present study, there is one button for moving forward in time and another for moving backward. Unlike the slidebar interface, where users need to focus on the sliding bar, button navigation requires users to tap the buttons. This interface design may draw less visual attention since the date and time are displayed separately from the buttons.

Finally, cyclical navigation, which is based on a circular control element, which users can interact with to select specific time periods. This interface differs from traditional linear timelines and offers an alternative approach to temporal navigation. Users can interact with the circular control element by tapping or dragging their finger across it. Circle navigation provides users with a unique way to explore spatiotemporal data. By selecting different segments or labels on the circle, users can choose specific moments.

All three navigation interfaces are presented in Fig. 1. For the cartographic background, we used administrative borders of Portugal. Forecast weather data include the maximum day (orange rectangle) and night temperature (blue rectangle), water temperature, and cloudiness. These are common meteorological characteristics provided by weather maps (Blok 2013). Time navigation allowed between the 18th and 29th of July.

Fig. 1
figure 1

Three experimental interfaces which enable time navigation in the study. Designed for a mobile device Samsung Galaxy M22: A slidebar navigation; B button navigation; C cyclical navigation. Letters (A, B, or C) were not visible during the experiment

Full size image

Participants

Fifty students of the Adam Mickiewicz University Poznan participated in the study experiment. They were divided into three groups. Each group had to perform a series of map-based task on the selected mobile map interface. The first group performed the experiment on the map with the slidebar navigation (A), the second group on the button navigation map (B), and the third group on the cyclical navigation map (C). All participants were asked about their gender and age. Participants were selected based on their experience in using weather apps on mobile phones and their familiarity with weather maps.

Each participant had prior experience with mobile and weather forecast maps. They were students from the geographical faculty, and they had a background in geography and cartography. They had normal or corrected-to-normal vision without astigmatism. Before the participation, informed consent was obtained from all participants. The institution in which the research was conducted did not require the permission of the ethics committee for conducting this study. Regardless of that, ethical guidelines and standards were followed. All participants’ characteristics are set in the Table 1.

Table 1 Study participants’ characteristics according to selected GUI
Full size table

Equipment

The Gazepoint GP3 HD eye tracker with a sampling frequency of 150 Hz was used in this study to record eye movements. The visual angle accuracy of the eye tracker was up to 0.5°, and the calibration procedure involved a 5-point method on the phone screen. The mobile phone used for displaying maps was a Samsung Galaxy M22 with a 6.4-in.screen. It was placed on a stand to avoid movement of the display. However, it allowed the device to be held in hand. The average distance between participants’ eyes and the eye tracker was approximately 58 cm ± 2 cm, and the participants’ average gaze sample score was 89% ± 1.7%.

Procedure

The experimental procedure started with oral introduction of participants and defining the purpose of the study. Participants were instructed about the color of daytime and nighttime temperatures, water temperatures, and how the interactions works. The next step was to perform a calibration with the eye tracker and perform a familiarization procedure, which involved free viewing and interactions with the given mobile map interface. Then, the participant was asked eight questions in a random order:

  • What was the highest water temperature, and when did it occur?

  • What was the lowest water temperature, and when did it occur?

  • What was the highest daytime temperature, and when did it occur?

  • What was the lowest daytime temperature, and when did it occur?

  • What was the highest nighttime temperature, and when did it occur?

  • What was the lowest nighttime temperature, and when did it occur?

  • What was the greatest temperature difference between daytime and nighttime, and when did it occur?

  • What was the smallest temperature difference between daytime and nighttime, and when did it occur?

Each question was asked separately (the full questionnaire of the study is attached in the Appendix). Participants gave oral answers to each question. The questions were designed to encompass all the information presented on the map, extending the map-based search task, and including additional comparison tasks, as represented by the temperature amplitude search. According to Andrienko et al. (2003), these types of tasks involve defining a single object within a single time unit as one of the four map-based tasks. The following procedure did not include spatial pattern or temporal trend searches.

Although the research employs eye-tracking to document user behavior during the evaluation phase, it does not explicitly investigate the application of eye-tracking during the design phase of the three interfaces.

Analysis

Firstly, we analyzed how many participants responded correctly to each question, and how many correct answers were given in total. Additionally, we considered time that each participant needed to provide the answer. This could be related to the effectiveness and efficiency measures (Çöltekin et al. 2009; Garlandini and Fabrikant 2009). The eye-tracking analysis was focused on fixation-based metrics and area of interest (AOI), especially, fixation location on the map display, fixation duration, total number of fixations, AOI viewed time, and AOI revisits. These metrics are considered markers of visual search in map reading process (Lloyd 1997, 2005), supporting users’ visual experience analysis (Cybulski and Horbiński 2020). Since our results did not have the normal distribution, we applied Kruskal-Wallis ANOVA for testing statistical significance.

Results

Participants’ performance when navigating time-based weather data on mobile maps

The study allowed us to find out that all three user interfaces were effective in detecting extreme water temperatures, with median effectiveness score of 91% for A, 97% for B, and 94% for C. The questions about maximal and minimal daily temperatures also had relatively high effectiveness scores, with the median score ranging from 74 to 78%. However, there were significant differences in effectiveness scores for detecting extreme night temperatures and highest and lowest temperature amplitudes, with median score ranging from 38 to 50% for night temperatures, and 44 to 65% for temperature amplitudes. Overall, the button navigation interface (B) was found to be the most effective GUI for weather maps. The detailed effectiveness results of all three map interfaces are presented in Fig. 2.

Fig. 2
figure 2

Detailed effectiveness results according to question type and GUI differences

Full size image

Also, the efficiency results were similar in the context of question type and GUI. Therefore, the Kruskal-Wallis ANOVA did not detect any statistical significance between different types of interface in terms of response time. The less efficient search included temperature amplitudes. However, searching for the highest amplitude was the most difficult task in every interface (A, 127 s; B, 131 s; C, 131 s; all Mdn). The most efficient search included searching for minimal and maximal day temperatures and water temperatures, which corresponds to the effectiveness results. Detailed results of response time (efficiency) are reported in Fig. 3.

Fig. 3
figure 3

Response time (efficiency) according to specific map-based tasks (questions). Abbreviations are: night.min/night.max, minimum/maximum night temperature; day.min/day.max, minimum/maximum day temperature; water.min/water.max, minimum/maximum water temperature; amp.min/amp.max, minimum/maximum temperature amplitude

Full size image

User experience implications of various temporal navigation interfaces

Participants interacted with different time navigation interfaces with the mobile weather map. In the case of slidebar navigation (A), participants could move the slidebar to the left and to the right smoothly. Most of them used one hand and moved the slidebar with a thumb even though the smartphone was placed on the stand. This type of interaction required moderate visual attention on the interactive slidebar since there were also dates associated with it. The median time of fixating the slidebar during map-based tasks on GUI A was 31 s. This corresponded with 6.1% of total time spent on viewing navigation interface (131 fixation revisits). As for the button navigation (B), most participants used both the hands and both the thumbs to move back and forward.

Since the date was displayed above the map content, participants rarely fixated on the navigation buttons. Therefore, the median time of fixations during map-based tasks on GUI B was 10 s. This corresponded with 1.9% of total time spent on viewing navigation buttons (51 fixations revisits). The circle time navigation (C) was the most fixation-consuming GUI, resulting in less map area fixations. Most participants used one hand, tapping the circle with the pointing finger. Since it was not a standard interface, and dates were associated to it, participants’ view time median was 59 s, which corresponded with 10% of total time spent on viewing navigation circle despite the map-based task (199 fixation revisits). Figure 4 presents user interface experience in terms of handling GUI and visual attention devoted to map area.

Fig. 4
figure 4

Different participants’ interactions with GUI. The percentage of map area and GUI referrers to visual attention based on median fixations. Kruskal-Wallis ANOVA results (H and p-value) of fixations on GUI among different tasks are reported at the bottom part of the figure

Full size image

Visual attention during map-based tasks on mobile weather maps

Fixation-based analysis includes Kruskal-Wallis ANOVA within-subjects facts, which were GUI type (3 conditions) and answer response correctness (2 conditions).

In the slidebar-type GUI (A), there was a statistically significant difference in the total number of fixations between successful and unsuccessful participants. It concerns searching for the lowest night temperature, and participants, who successfully found this value on the map, had significantly higher number of fixations (215 vs 135) during this task (H = 6.25, p < 0.01). The second fixation-based metric that was significantly different among successful and unsuccessful participants was the fixation duration. In the task in which people searched for the highest day temperature, the successful participant had a longer single fixation than the unsuccessful one (median of 280 ms vs 180 ms).

In the button-type GUI (B), there was a significant difference in the total number of fixations among successful and unsuccessful participants while searching for the lowest and the highest temperature amplitudes (H=5.415, p<0.02 for max. amplitude, and H=4.88, p<0.027 for min. amplitude). During these activities, participants, who were successful in finding the lowest and the highest temperature amplitudes, had significantly more fixations on the map (more than 100 fixations per task). In terms of saccadic magnitude, successful participants had significantly shorter distances between fixations (median 93° vs 174°). However, this was only significant during searching for the highest temperature amplitude (H=4.01, p<0.04).

In the circle-type GUI (C), we did not find any statistically significant differences between fixation-based metrics.

Figure 5 presents how visual attention (based on fixation count) is distributed on the map, according to each type of map-based task. In some cases, there are clear differences between button-type GUI (B) and other interfaces. When GUI A and C present similar patterns of visual attention, GUI B shows a significantly different way of fixation distribution (e.g., searching for the highest and the lowest night temperatures).

Fig. 5
figure 5

Participants’ visual attention on the map according to each map-based task and GUI type. The more red the more visual attention (based on fixations) in this part of the map

Full size image

In tasks in which participants had to find the highest and the lowest water temperatures, it can be noticed that visual attention is focused on the left side of the map content, where the water temperatures are presented. Compared to effectiveness results in terms of searching for day temperatures, the GUI C showed lower effectiveness than other interfaces. The heat map analysis shows that in this case, the visual attention is not focused in one area as much as in GUI A and B. Therefore, participants experienced more difficulties in visual search while performing with the circle-type GUI.

The highest differences in effectiveness were visible during the task of searching for the highest and the lowest temperature amplitudes. The GUI B gives the best results in this matter. Heat maps show great differences in visual attention distribution among tested GUIs. While there are some similarities between GUI C and B while searching for the lowest temperature amplitude (the main focus point being shifted), there are no such differences while searching for the highest temperature amplitude. On the other hand, there are similarities between GUI A and B. Therefore, searching for the same information but on a different GUI can lead to some differences in visual attention, and this can affect effectiveness and efficiency results.

Discussion

The results suggest that the layout of the GUI may have a significant impact on the user’s visual strategy and their ability to effectively and efficiently complete tasks. This corresponds with the Harrower and Fabrikant’s (2008) research on different GUI types. The effectiveness results show that GUI B was the most effective out of three interfaces, with higher median effectiveness scores in almost all tasks. This indicates that the layout of GUI B, which included a slidebar for temporal navigation and a heatmap for displaying temperature data, may have facilitated the users’ visual strategy in a way that allowed them to detect patterns in the data more easily and accurately.

On the other hand, GUI A and C had lower median effectiveness scores, particularly in tasks related to detecting extreme night temperatures and highest and lowest temperature amplitudes. This suggests that the layout of these interfaces may have hindered the users’ visual strategy in some way, making it more difficult for them to perceive and interpret the data accurately.

Based on the study results, users perceive and interact with different GUI designs differently while navigating through time-based weather data on mobile devices. The research study by Horbiński et al. (2020) focuses on the location of specific buttons on mobile devices, such as geolocation, zoom, search, and route find. The study suggests that mobile devices require a different way of handling interactive activities than desktop solutions, and that interactions on mobile devices should be placed differently to cater to users’ preferences and habits. Our study showed that, on mobile devices, handling is related to different temporal navigation interfaces and should be considered as a part of map design, especially when exploring weather maps.

The study showed that different temporal navigation interfaces (GUI designs) led to differences in visual attention and performance outcomes when participants were tasked with finding specific weather data on the mobile weather map. The study allowed us to find out that certain GUI designs were more effective and efficient for certain tasks while others were less effective and efficient. This corresponds with the study by Bartling et al. (2021b) which suggests that GUI design supports effectiveness and participants’ self-confidence. Therefore, it is important to consider GUI design when designing map-based tasks for mobile devices to optimize performance outcomes. The comparison between small multiple maps and interactive animation on mobile devices is a topic of interest in cartography and user experience design (Harrower and Fabrikant 2008).

Both approaches have their advantages and disadvantages, and the choice between them often depends on the specific goals of the map and the preferences of the target audience. Small multiple maps, also known as “small multiples” (Griffin et al. 2006), provide a clear and structured presentation of information, making them suitable for tasks that require careful examination of individual maps (Gołębiowska et al. 2020). On the other hand, interactive animation on mobile devices allows users to explore temporal and spatial changes in data dynamically. This approach can be engaging and visually appealing, as it provides a sense of continuity and allows users to see how data evolves over time (Cybulski 2021). In summary, the choice between small multiple maps and interactive animation on mobile devices should be guided by user needs, data complexity, preferences, device constraints, and the context of use.

There are several limitations to this study that should be noted. Firstly, the sample size was relatively small, consisting of only 50 participants. This may limit the generalizability of the results. We acknowledge that restricting the study participants to a specific group with shared geographical and cartographic backgrounds, similar age profiles, and the absence of visual impairments may limit the generalizability of our findings to the broader population. Secondly, the study only focused on a specific type of map-based tasks related to weather data exploration, which may not be representative of other types of map-based tasks. Thirdly, the study failed to take individual differences in mobile device usage and familiarity into account, which may have affected participants’ performance and preferences. Finally, the study did not investigate the long-term effects of different GUI designs on user experience and performance, which could be a valuable area for future research.

Based on the limitations of this study, there are several directions that future research could take to expand on our findings. Some potential paths for future research include:

  1. 1.

    Investigation of the effect of different GUI designs on other types of map-based tasks beyond the weather data exploration tasks used in this study.

  2. 2.

    Examination of the impact of GUI design on the performance of different user groups, such as individuals with varying levels of technological literacy or individuals with visual impairments.

  3. 3.

    Investigation of the impact of GUI design on user satisfaction, as well as other subjective metrics, such as mental workload and user engagement.

  4. 4.

    Exploration of the impact of different temporal navigation interfaces on other types of mobile applications beyond map-based tasks to understand the generalizability of our findings.

Conclusions

It is important to consider users’ experience in the development of cartographic products, as users have individual preferences that may affect their performance. While effectiveness and efficiency studies can provide valuable information, visual and interaction experience may reveal additional insights that may be crucial for understanding users’ preferences and performance. The example of the scrolling mechanism on social media and its role in addictive behavior highlights the importance of considering the psychological effects of design choices (Burhan and Moradzadeh 2020; Baym et al. 2020). Similarly, the study carried out on mobile weather maps demonstrates that the design of the user interface may have a significant impact on users’ visual attention and performance.

The finding that the separation of the date from the time navigation panel reduces necessary visual attention is a valuable insight for future GUI design. The button interface is also shown to be effective in supporting visual attention on the map itself, which is important for tasks, such as searching for specific information. Overall, it is important to consider both the functional and experimental aspects of cartographic product design to create products that are effective, efficient, and enjoyable for users. Further research can continue to explore different design choices and their impact on users’ experience and performance.

It is important to note that each GUI design has its own unique features that may be useful for different mapping solutions. While the GUI button was found to be effective in supporting visual attention on the map and separating display date from interactive elements, other GUI designs may have their own strengths and weaknesses. Table 2 can provide valuable information on the unique characteristics of each tested GUI, which can be helpful in determining which design is the most suitable for a specific mapping solution. It is important to consider specific needs and preferences of the target user group when selecting a GUI design, as this may have a significant impact on their experience and performance with the mapping product.

Table 2 Unique characteristics of tested GUI according to the study results. (1) Handling was limited due to the necessity of holding the smartphone steady on the cradle
Full size table

Overall, it is important to carefully evaluate and compare different GUI designs to determine the most effective and efficient solution for a specific mapping application. This may involve conducting user testing and gathering feedback to ensure that the design meets the needs and expectations of the target user group. The findings of this study contribute to the growing body of research on mobile cartography and provide practical implications for designers and developers of mobile weather applications. Further research is needed to explore the impact of other factors, such as users’ prior experience and cognitive load, on the performance of map-based tasks on mobile devices.