1. Introduction
In 2009, the second-largest city in the Kingdom of Saudi Arabia (KSA), Jeddah, experienced devastating rainfall and flooding, leading to approximately 122 deaths, 350 missing individuals, and $3 billion (U.S. dollars) in damages (Youssef et al. 2016). In 2015, another deadly weather event occurred in Mecca’s holy mosque when a large crane collapsed due to heavy rain and damaging winds killing more than 107 people and leaving 230 people injured (Gul and Ali 2016). These unfortunate incidents could have been alleviated if sufficient and timely weather information, including details about possible impact and consequences, were provided. These life-threatening and time-critical weather conditions have increased impact on perhaps most Middle Eastern countries due to several reasons, including low investment in weather forecasting services, poor infrastructure, and lack of compliance to safety standards and regulations. With the dynamic nature of weather, accessing weather forecasts and warning alert notifications is critical for safety. This is even more crucial during seasons known to have destructive weather conditions (e.g., floods and sandstorms) that require frequent accessibility as well as prompt and appropriate reaction by the public.
The Saudi Vision 2030, established in 2016, has included several development initiatives concerning all governmental and private sectors. The main goals are to enhance all public services provided to Saudis and residents, as well as create different income sources rather than depending solely on oil (Al Arabia 2016). One of the Saudi initiatives is establishing several weather forecasting centers supplied with the latest technologies and advanced forecasting equipment and tools to provide possible accurate forecasts and warnings.
Enhancing weather forecasting services’ technological aspects is critical for the accuracy of broadcasted information and forecast lead time. However, focusing only on the technology while initially neglecting how people perceive weather forecasts or what they need will likely result in unsuccessful final products (Chan-Olmsted et al. 2013). It is fundamental to learn about the perceived value of weather forecasts among the public, the most important and critical weather features, and the availability of such features in existing popular weather information delivery sources, such as smartphone weather apps (Demuth et al. 2011). Learning about these factors may significantly help devote the knowledge acquired through advanced technologies toward the public’s actual needs. To the best of our knowledge, no work has been done on understanding how the public thinks about the provided weather services in KSA or what they need.
2. Background
a. People’s perception of weather forecasts
Numerous governmental and private agencies oversee weather phenomena and promptly disseminate various weather forecasts, ranging from normal weather conditions to severe weather warning alerts. These weather forecasts aim to help the public make informed decisions for planning their daily activities and, more importantly, saving their lives and properties (Cools and Creemers 2013). Several peer-reviewed articles examined how the public perceives, responds, and uses the various issued weather data (Silver 2015). Conducting studies that pertain directly to specific nations with regard to their perception and response to various weather conditions is important. Several factors (e.g., traditional knowledge, prior experience with hazards, and social demographics) can greatly influence the public’s perception and reaction (Wu et al. 2017). For instance, Drost (2013) and Hayden et al. (2007) investigated American citizens’ perceptions of tornado and flood warnings, respectively. Both studies indicated the Americans’ high-level perception of severe weather conditions and their need for detailed, clear, and timely information, including precautionary advice, to precisely determine the risk level associated with the severe weather condition and ultimately make informed safety decisions. Similar results were attained in Wong and Yan’s (2002) study, which were perceived by the Chinese public. Silver and Conrad’s (2010) study indicated a similar high-level perception of severe weather conditions among Canadians when compared with Americans and Chinese. It was concluded that weather data needed to be tailored based on the different weather delivery means used by target groups. In addition to the past research on people’s perception of severe weather conditions, Lazo et al. (2009) and Demuth et al. (2011) determined Americans’ perceptions, uses, and preferences of nonsevere weather conditions. Both articles showed that Americans’ perceived basic weather information is important for daily life activities and revealed large discrepancies in people’s attitudes and behaviors toward such information. Findings from these studies have been utilized for optimizing the communication between weather agencies and end users (Zabini 2016).
Measuring the specific individuals who will be receiving weather forecasts, alerts, and risk communications is especially pertinent as it ultimately influences the protective action decision-making and behavioral response of the individual [see the protective action decision model (PADM) by Lindell and Perry 2012]. This model suggests that constructing emergency response messages should be “personally relevant, appropriate, and focused on personal efficacy” (Heath et al. 2018). Researching environmental cues, social cues, information sources, channel access, and preferences, warning messages, and receiver characteristics are all imperative according to PADM to construct appropriate weather alerts and risk communication but to also begin building a foundation for a large-scale weather infrastructure as seen with implications in Lindell and Perry (2012). As there have been no previous efforts to look at how this model can be applied to KSA residents, this study is meant to explore the preliminary application of PADM to non-Western individuals.
Although these research articles present valuable insights on how the public perceives and responds to weather forecasts, most of the work has been done from a Western perspective. When benchmarked with Western countries, there is a significant lack of research on the human characteristics in accordance with life-saving data (i.e., weather forecasts) in KSA. In addition, no published research has been found on the public’s perception of either time-critical or non-time-critical weather forecasts in Saudi Arabia or any Middle Eastern country. Saudi Arabia represents nearly 80% of the Arabian Peninsula (Darfaoui and Assiri 2010) and has very different weather features from Western countries, as it is geographically located in the desert. For these reasons, Saudis’ perception of weather forecasts and the accessibility importance of the means through which these weather forecasts are delivered are major concerns in this research.
b. People’s most required weather features/information
Investigating what type of weather information people are mostly concerned with and the reasons for using specific weather features is important for weather agencies to enhance their final weather products and essentially meet people’s needs. For example, Lazo et al. (2009) found that basic weather forecasts (e.g., temperature, precipitation, and wind speed), in addition to location services are perceived to be valuable and sufficient information through weather delivery means. Another study done by Khamaj et al. (2019) interviewed a sample of U.S. residents and revealed that detailed weather forecasts, including alert notifications of extreme weather conditions, are required. The study further illustrated that people might base their belief of required information on several factors or situations such as the nature of climate and geographical aspect of their area and the type of activities they may do in everyday life (Khamaj et al. 2019). In line with Khamaj et al.’s (2019) findings, Weiss and Wobb (1986) surveyed several farmers by asking mainly about the most important weather information for their agriculture jobs. The respondents showed a major reliance on short-term weather forecasts, focusing on rainfall levels (Weiss and Wobb 1986).
However, with advanced meteorological tools weather agencies currently have and people’s major reliance on weather data for planning daily activities, people may tend to look for detailed weather information characterized under multiple different features. Several weather providers, especially through modern means (e.g., websites and smartphones), enable users to access such information in both textual and graphic formats. With such, users can add locations of interest, receive watch and warning alert notifications, access forecasts on maps, and even tailor and customize the settings to show the type and amount of needed information. Several studies (e.g., those by Khamaj and Kang 2018; Jon et al. 2018; Wu et al. 2015; Phan et al. 2018; Drogalis et al. 2015) have investigated people’s perceptions of such features in terms of the necessity to certain areas, type of individuals’ activities, and specific weather conditions. Moreover, studies by Ash et al. (2014) and Sutton and Fischer (2021) determined people’s attention and responses to severe weather data varied when the color schemes of the visual displays were altered. The findings from both studies greatly support the need for using visualization techniques to enhance the public’s processing of potential weather hazards and efficiently make informed safety decisions. However, the results suggest considering the variant characteristics of the public when designing graphical displays. KSA is currently in the initial steps toward developing its meteorological operational aspects, and therefore, it is crucial to learn about the public’s required weather information and prioritized features.
c. Smartphones and weather
People can utilize several popular sources for information, such as television, websites, and radio. For instance, a study by Tsai et al. (2018) showed data that many people flocked to social media to help understand and share their experiences during natural disasters (e.g., a Taiwanese earthquake). However, with the rapid progression of smartphone technology and the convenience of accessing the information on smartphones at any time and location, people worldwide tend to rely on smartphones more than other information sources (Nayebi et al. 2012). As an example of the magnitude of smartphone usage, about 83% of Saudi adults (age 18+) own and use smartphones for online information (Statista 2019a). The large ownership and usage of smartphones have led businesses and technology experts to develop various downloadable applications (apps) continuously. While no published statistics have been found on smartphone app downloads and usage in Saudi Arabia, a large number is estimated. This estimation is supported by the large percentage of Saudi smartphone users and the fact that one of the top goals of using smartphones is to utilize continuous information on categories such as games, weather, news, and sports through downloadable apps (Purcell 2011).
With the increasing use of smartphone apps globally, the weather category’s apps are becoming among the top downloadable app categories (Statista 2019b). For instance, The Weather Channel app has been the top downloaded weather app, with more than 7 million annual downloads in the iTunes store alone (Sensor Tower 2020). Regardless of the context of use, accessing immediate life-threatening information, such as that of weather, and performing related life-saving actions can be easier and quicker on smartphone apps than any other sources (Gökçearslan et al. 2016). However, the type and amount of weather information have been of major concern. With the small screen size of smartphones, the large amount of information meteorological authorities deliver to weather app businesses, and the life-threatening nature of time-critical weather incidents, it may be crucial to learn about end users’ specific needs to enhance the presented content (Khamaj et al. 2019). The information collected from end users with regard to their nature of climate and type of outdoor activities may greatly help both users access the required information and app developers in growing their businesses. In addition, following the PADM model considering the public’s different characteristics (e.g., in terms of age, education, and ability) and their varying needs when creating and broadcasting weather forecasts, especially warning messages, could be imperative for understanding the weather phenomena and make timely decisions (Bean et al. 2015; Lindell and Perry 2012).
d. Study research questions
Overall, this study aims to fill the gaps found in previous research and act as preliminary research in developing services for the public by addressing the following operational research questions:
To understand what future risk communications would be most effective, what is the perceived value of weather forecasts among the public in KSA?
What weather features and information the public in KSA perceive as important or required?
Do existing smartphone weather applications in KSA provide users with the required features and information?
3. Methods
a. Overview
The study employed a cross-sectional questionnaire via the Qualtrics software to investigate people’s perception, and behavior of weather features and delivery means in KSA. Specifically, the questionnaire surveyed Saudi citizens and residents at all KSA’s regional areas (see Fig. 1) in terms of their demographics, practices toward current weather forecasts and delivery means, the belief of the importance level of weather forecasts accessibility, perception of critical weather features, and experience with the available features on existing smartphone weather apps in KSA.
A map of KSA showing the study area (center, south, north, east, west) surrounded by red circles; the thicker the circle border is, the more respondents are recruited from that area. KSA represents the central part of the map, with other countries and the Arabian Gulf surrounding it from the south, east, and north. The Red Sea surrounds it from the west (source: University of Texas Libraries; http://legacy.lib.utexas.edu/maps/middle_east_and_asia/saudi_arabia_admin-2013.pdf).
Citation: Weather, Climate, and Society 13, 3; 10.1175/WCAS-D-20-0152.1
b. Participants’ demographic and recruitments
The study included a convenience sample of 315 participants. Of the 315 participants, a majority were male (65%), were aged mostly between 20 and 49 (89%), were educated, with a majority reporting holding a bachelor’s degree (47%) and were from primarily the east region of KSA (56%) and the south region (35%). The target population of this study was both citizens and foreign residents in KSA. The participants were included in the study if they consented to participate. To maintain some form of representation, the study recruited participants of different demographics such as age, gender, education levels, regions, and weather experiences; the data on participant demographics can be seen in Table 1). Recruitment was done between March and June 2020 by distributing the Qualtrics questionnaire link through social media and emails. The questionnaire link with a brief message was initially sent out to random celebrities and public figures within the different geographical regions of KSA, who agreed to help in the research. This initial recruitment strategy was performed by searching for popular accounts within social media using keywords such as region names, well-known family names within each region, and meteorology accounts. They then forwarded the invitation requests for participation to their followers and lists of beneficiaries. The message associated with the questionnaire link also encouraged respondents to share the link with other fellow KSA residents. No direct contact existed with the participants, but they completed the questionnaire electronically once they received and opened the link. The questionnaire was in Arabic (most respondents’ mother tongue) and English (minorities’ first language); the participant can select their preferred language as soon as they open the link.
Respondents’ demographic and general weather accessibility practices.
c. Questionnaire structure
The questionnaire consisted of 30 questions categorized under four main themes based on the PADM. The questionnaire solicited demographic information (i.e., age, gender, education level, and city of residence). Following these questions, participants were asked about weather forecasts and delivery means in general, specifically to what extent they are familiar with weather products provided by the Saudi Presidency of Meteorology and Environment (PME), their preference of weather delivery means, their frequency rate of accessing weather forecasts, their current most used weather delivery means, and whether their city experienced extreme weather conditions during the last five years. The next set of questions asked respondents about the importance of accessing weather forecasts by explicitly asking about the role of forecasts in saving people’s lives and properties and planning daily activities such as outdoor events and clothing decisions. Then, the question shifted to ask about respondents’ perception of required weather features (i.e., detailed weather forecasts, exact location search, warning alert notifications, detailed alert messages, radar maps, road forecasts, and forecast customization) on weather delivery means. Respondents were finally prompted to determine whether their required features are available on current smartphone weather apps in Saudi Arabia. Most questions featured a 5-point Likert scale (1 = strongly agree, 2 = agree, 3 = neither agree nor disagree, 4 = disagree, and 5 = strongly disagree; note: the last set of questions included 6 = I do not know). Other survey questions consisted of multiple-choice items where participants selected one answer from a list. The multiple-choice questions included participants’ demographics such as age, gender, region, and education level. Also, multiple-choice items about general weather accessibility information were included, such as preferred weather delivery means and frequency of checking weather forecast information. Please see Table 1 for details.
d. Pilot test
Forty participants (about 13% of the final sample) took part in the study to pilot test the questionnaire. The responses of the pilot test were excluded from the study to account for response bias. The pilot test was used to examine the questions and response items in terms of comprehension. The pilot test results guided the researchers to make several modifications to question wording and response items.
e. Data analysis
Both descriptive and inferential analyses were performed on the dataset to test the research questions and investigate relative variables. Specifically, descriptive statistics were utilized to address all three research questions of the study. One-way ANOVA was performed to see group differences concerning the importance of weather forecasts, amount of information, and the need for adding new features to weather apps. Two stepwise linear regression tests were conducted to 1) predict whether KSA residents perceive weather forecasts to not be important and 2) investigate perceptions of if weather awareness saves lives. To measure the strength and direction of association between several key variables, correlation tests were administered. In addition, reliability tests were conducted to measure the internal consistency of the survey items.
4. Results
The results below are mostly descriptive, preliminary analyses showing some KSA citizens’ perceptions on the importance of having a weather application, what features these individuals would like to have in a weather app, and data on what weather applications currently have.
a. General weather accessibility information
As shown in Table 1, most participants stated they were familiar with available weather applications (54%). According to these data, individuals checked the weather application only when needed (34%), followed by 2–5 times a week (26%). The data suggest that individuals used smartphone applications to check the weather (68%), and 69% thought it was the best form of media to check the weather. To summarize these data, individuals often check weather information when it is relevant via primarily smartphone devices in KSA. How about for the individuals that do not find weather forecasting important? Interitem reliability was tested between the three questions asking, “weather forecasts are NOT important” and “Nothing is necessary for weather applications,” finding a Cronbach’s alpha of 0.64 indicating some consistency between the measures, meaning that some participants answer the two questions similarly. The interitem correlation [r(315) = 0.47; p < 0.001] seems to also indicate that if you believe weather forecasting is not important, then an individual is likely to also believe that nothing is necessary in terms of features.
b. Implication of weather forecasts and its perceived value
The results in Table 2 show comparisons on KSA citizen perceptions on the importance of being alerted about the weather (especially severe weather) with regard to daily activities (e.g., what to wear), values (e.g., saving lives), and if they do not think weather alerts are important. There is acceptable internal consistency (Cronbach’s alpha = 0.77) for the questions relating to weather forecasts’ perceived value. It is good to report the alpha values for new scales or measures as it indicates internal consistency. Generally, when looking at differences between questions (variance), it is necessary to know if the relationships are due to the measured construct rather than measurement error, which Cronbach’s alpha estimates (Tavakol and Dennick 2011). Overall, individuals in this sample disagree on thinking the weather is not important and seem to believe that weather alerts impact daily activities and values. This, with the demographic information, would imply that perhaps KSA citizens believe that being alerted to weather events is important and impacts their lives and would like this information via smartphone.
Respondents’ perceptions of the importance level of weather forecasts accessibility (n = 315).
c. Importance of some features of weather forecasts
The results in Table 3 show KSA citizen perceptions on the necessity of various weather functions. There is good internal consistency (Cronbach’s alpha = 0.85) for questions relating to weather forecast features’ perceived importance. Overall, individuals seem to agree that all listed functions are necessary for a weather delivery source. These data could be used to inform future and current development of weather communication technologies. Further, it may help plan important weather risk communication for when events do happen.
Respondents’ perceptions of crucial weather forecasting features (n = 315).
d. Existence of common weather forecasting features on smartphone weather applications
The results in Table 4 show KSA citizen perceptions on the availability of various weather application functions on current weather applications. There is good internal consistency (Cronbach’s alpha = 0.85) for questions relating to weather forecast application features. There seems to be more disagreement on the availability of current application features, which can be seen in the percentage of individuals who marked “NA” or “do not know.” Further, the percentage of individuals who agree or disagree are relatively close on some of the questions. This can be seen visually in Fig. 2. Reasons for the disagreement may be a function of interactivity with the current apps and the apps themselves. Future research is needed to investigate what specifically leads to an individual knowing what features are detailed in a weather app.
Respondents’ experience with smartphone weather applications in KSA (n = 315).
A bar chart of what features individuals think are part of weather applications currently. Each category is the percent of individuals that agree, disagree, or do not know if that feature is currently on weather applications.
Citation: Weather, Climate, and Society 13, 3; 10.1175/WCAS-D-20-0152.1
e. Importance versus existence of common weather forecasting features
To understand which app weather feature is more important based on the responses received from the participants, a priority index (see Table 5) has been developed. The index is calculated by dividing the percentage of participants who agree that the feature is important by the percentage of those who agree that the feature exists. The weather app developer can utilize this index to prioritize the features that they need to work on next since it represents highly needed features that do not exist. Looking at the computed indices, a weather app developer may choose to work on the feature that allows the users to receive weather forecasts for exact location (the first item in Table 5 with the highest priority index of 3.04). For the detailed forecasts feature (priority index of 1.47), 86% of the participants indicated that it is important, but 58% indicated that it exists.
Priority index description of weather forecasting features. The priority index = percent of agreement of importance/percent of agreement of existence.
f. Correlation
The positive correlations found in Table 6 overall seem to indicate that KSA individuals in this sample generally think weather awareness saves lives or property tend to want weather alerts and forecasts to their exact location. Further, those who think that weather impacts their daily decisions (e.g., outdoor activities, clothing choices) tend to want customizable controls in their weather app [r(315) = 0.34; p < 0.001]. Single main effects of age see a very small positive correlation with belief that weather saves property [r(315) = 0.12; p = 0.04] and desire for a detailed weather forecast [r(315) = 0.12; p = 0.03]. There is a strong correlation between the judgment that weather impacts outdoor activities with judgments that weather impacts clothing decisions [r(315) = 0.62; p < 0.001]. Further, if an individual does not think that weather is important, they are likely not to judge that weather impacts outdoor activities [r(315) = −0.37; p < 0.001] or clothing decisions [r(315) = −0.32; p < 0.001] or that weather forecasts save lives [r(315) = −0.35; p < 0.001] or property [r(315) = −0.32; p < 0.001]. A positive correlation can be seen between those who answered weather is not important and thinking that nothing is necessary for the weather application [r(315) = 0.47; p < 0.001]. The lack of significant correlations for current application use and age may indicate that the wanting of a weather app may stem more from beliefs about the weather (e.g., weather awareness saves lives).
Correlations of selected weather variables. Positive correlation with age implies older individuals. Gender was coded as 1= male and 2 = female. One and two asterisks indicate statistical significance at p < 0.05 and p < 0.001, respectively.
g. Demographic differences
Significant demographic correlations saw small correlations with females believing weather forecasting is not important [r(315) = 0.11; p = 0.043] and believing that nothing needs to be added to weather apps [r(315) = 0.12; p = 0.031], while males saw a small correlation with wanting a detailed weather forecast [r(315) = −0.14; p = 0.013]. There were no significant correlational differences between all other variables and gender. One-way ANOVA results concurred with the correlational results finding that males and females significantly differed in their opinions of believing weather is not important [F(1, 314) = 4.11; p = 0.043], wanting a detailed weather forecast [F(1, 314) = 6.25; p = 0.013], and believing nothing is necessary [F(1, 314) = 4.72; p = 0.031]. Significant, small correlations saw that older individuals tended to believe that weather awareness saves property [r(315) = 0.12; p = 0.035] and wanted a detailed weather forecast [r(315) = 0.12; p = 0.029]. All other correlations with age were not significant.
h. Select regression analyses
A stepwise linear regression was used to predict if a KSA resident may perceive weather forecasts to not be important. Table 7 shows the stepwise model. The final model shows that the best predictor of if KSA residents may perceive forecasts not to be important if they believe there is nothing necessary to be added to weather applications [R2 = 0.22, β = 0.47, t(314) = 9.36, and p < 0.001]. Next, believing that weather forecasts do not impact outdoor activities and believe that weather awareness does not saves lives, predict thinking that weather forecasts are not important added 0.096 of explained variance. Finally, being less educated is associated with thinking that weather forecasts are not important added 0.013 explained variance for a final model with a R2 of 0.319.
Stepwise regression for predicting whether individuals believe weather forecasts are not important. One and two asterisks indicate statistical significance at p < 0.05 and p < 0.01, respectively.
A second stepwise linear regression was used to investigate perceptions of if weather awareness saves lives (Table 8). The top predictor of if KSA residents believed if weather forecasts save lives is if the individuals also believe if weather forecasts save property [R2 = 0.26, β = 0.52, t(313) = 10.62, and p < 0.001]. The following predictor is believing that weather impacts outdoor activities (R2 = 0.05), and the next is believing that weather alerts are required (R2 = 0.02).
Stepwise regression for predicting if individuals believe weather awareness saves lives. One and two asterisks indicate statistical significance at p < 0.05 and p < 0.01, respectively.
5. Discussion
a. Main findings
Research has investigated people’s perceptions and attitudes toward weather services provided in Western countries. However, to the best of our knowledge, no published work has pertained directly to understanding the various facets of what and how people perceive the current weather services offered in KSA or any other country in the Middle East. Due to the representation of KSA to nearly 80% of the Arabian Peninsula, its unique geographical and weather aspects, and its promising future in different sectors, including the meteorological sector, this study aimed to mainly explore how to apply the PADM with Saudis’ perception of 1) the impact of weather forecasts on their everyday activities and values (receiver characteristics), 2) the necessary weather forecast features (information sources/channel access and preference/social cues), and 3) the availability of their required weather forecast features on smartphone weather apps (warning message preferences).
The study found that participants’ views on the importance of accessing weather forecasts relate to majorly enhanced awareness of the impact of weather data on people’s lives and everyday activities (see Table 2). These data implied that people who care about the weather might integrate forecast information into their daily decisions, like their choice of clothing or outdoor activity. The results also indicated that regardless of demographics, Saudis may need to know the individual effects of weather forecast conditions (both time-critical and non-time-critical conditions) to plan accordingly. The results further suggested that Saudi Citizens, contingent on the sample being representative, tend to want improved weather awareness and demand weather services that must be acted upon by both Saudi governmental and private meteorological agencies. Improving meteorological operational tools and ways of communicating weather data to end users may lead to improved safety for individuals and property while also delivering an improved user experience that may improve weather forecasts delivery means’ interaction.
The results about desired weather features revealed that most participants perceived all listed features as necessary on weather information delivery means, emphasizing warning alerts, alert messages for exact locations, and weather forecasts for specific road routes (see Table 3). These findings indicate that Saudis are aware of the importance of accessing weather forecasts and require detailed information about the weather phenomena and the impact that weather conditions may bring. The higher interest in obtaining information through the three features mentioned earlier is in line with some of Khamaj and Kang’s (2018) findings, which imply that people are mostly concerned about the features containing time-critical and life-threatening data. Specifically, receiving and accessing warning alert notifications of severe weather conditions (e.g., floods, tornados, or sandstorms), including information such as areas under alert, alert start/expiration time, possible impact, and precautionary advice, may greatly help people to understand forecast conditions and plan life-saving actions. Comparing this with the findings by Lazo et al. (2009) and Khamaj et al. (2019), Americans’ most-used weather forecast features are related to when and where severe conditions will occur. Lazo et al. (2009) and Khamaj et al. (2019) argued that these features were among the most used features because “people’s activities are affected by the weather at specific places and times” (p. 791). Lazo et al.’s (2009) and Khamaj et al.’s (2019) arguments are aligned with this study’s findings, where respondents indicated the importance of having weather forecasts tied to their exact location. Accessing weather forecasts for specific road routes may help alert people about possible extreme weather situations during travel (e.g., by car), mainly because KSA and many other Middle Eastern countries still have weak road infrastructure. This indication is supported by several tragic incidents of people dying after submerging their vehicles in flood water due to the lack of suitable drainage systems (FloodList 2019). Beyond saving lives, weather alert systems may also have economic impacts. One monetary example of an early flood warning system in Europe showed a savings of EUR 400 (USD ~450) for every EUR 1 (USD ~1.13) invested (Pappenberger et al. 2015). Weather alert systems with enhanced functionality and usability would also greatly ease users’ cognitive processes and decision-making strategies (Khamaj et al. 2019), enable users to do work-related tasks effectively and efficiently (Argyle et al. 2017), obtain user satisfaction (Rianti et al. 2020), leading to ultimately successful forecast products for both users and weather organizations.
The findings of this study further support the statistics about Saudis’ widespread adoption of the smartphone technology, as nearly 70% (see Table 1) of respondents believe that smartphone apps are the best means for communicating weather information to the public, and about the same percentage of respondents have been actively using them. These findings also support the literature on the magnitude of smartphone weather app usage worldwide because of its convenience and usefulness in different contexts of uses in comparison with other sources (Phan et al. 2018; Bryant et al. 2015). Even though more respondents agreed that existing weather apps contain the required features than disagreed, a considerable number of the respondents either did not know what features their weather app contained or were neutral on the topic (see Table 4). Looking at the results that “disagree” with the notion that current weather apps have good enough features and the “do not know” results suggested that current weather apps lack necessary features or may not invite users to explore the features of the current apps. Considering the initial stage of the user-centered design approach: determining user requirements, which was employed in Khamaj et al.’s (2019) study, smartphone weather app developers in KSA may need to enhance their systems based on users’ feedback in the current study, with giving more priority to the features listed in the priority index table (see Table 5). Our results indicated that enabling users to access weather forecasts and receive alerts for specific locations (e.g., house, school, and hospital) should be prioritized when designing or enhancing weather apps, as this feature scored the highest index rate. This might be attributed to the frequent false alarms associated with the current general location search methods (e.g., city and state or county). Specifically, with general saved locations, apps would still send warning alerts even if the user’s exact location does not fall within the issued geographical warning box by the PME. Further research is needed to explore how Saudi citizens perceive extreme weather alerts and current weather applications to better communicate risks and design the applications.
The positive correlations found in Table 6 seem to indicate that people who generally think weather awareness saves lives or property tend to want weather alerts and forecasts to their exact location, thinks that weather impacts their daily decisions (e.g., outdoor activities, clothing choices), and want customizable controls in their weather app. These results supported the results from the stepwise regressions in Tables 7 and 8, indicating that the best predictor of thinking the weather is not important is believing that current weather applications do not need more features. Further, if an individual does not think that weather is important, then they are likely not to think, given the negative correlations and regressions, that weather impacts their daily decisions or that weather awareness saves lives or property. The correlations and regression models indicated that wanting a weather app might stem from beliefs about the weather (e.g., weather awareness saves lives).
b. Limitations and future research
The results may not be completely representative and generalizable to the whole KSA population, but they produced a valuable initial insight from individuals varying in age, gender, educational levels, regions, and weather experiences toward building extreme weather forecasting communication technology in KSA. In other words, as this is the first effort or study related to weather attitudes of KSA residents, this method was used to look at initial exploratory outcomes that can be applied to a more probabilistic sampling method.
Ninety-one percent of the survey responses were obtained from people who reside either in the east or the south province of KSA. Given that different KSA regions are affected by different weather phenomena and the various subtle cultural differences, the opinions expressed in the current study may not reflect the general opinion with regard to the investigated research questions through the survey. A larger and more representative sample that proportionally covers all different regions of KSA is necessary. Some of the potential venues that can be utilized in future studies to achieve such coverage while ensuring a high response rate are to collaborate with local authorities such as weather forecast centers and universities and local news channels to promote the survey and increase its reach.
Currently, the study only utilized the survey research approach to investigate the citizens of KSA about their needs, opinions, and usage of the local weather forecast services. However, to ensure that these needs, opinions, and usages expressed in the survey are genuine, other research methods such as interviews, focus group studies, and diary studies are necessary. By utilizing different methods and triangulating the results, authorities in charge of making logistic changes and weather app developers will buy in and support pursuing research and app development in the direction of the findings. Further, according to the PADM model, exploring the target audiences of future risk communications is important. Utilizing other research methods, especially those that involve direct contact with the participants, such as interviews and focus groups, will ensure more understanding of the motive behind any need or opinion, which is typically hard to capture through surveys.
Further, understanding representative human behavior during a crisis may set the stage for future analyses that help with emergency evacuations and preparedness, like the agent-based simulation study on nuclear evacuation by Na and Lee (2016), which helped identify the predicted best plan of evacuation under different weather and wind conditions. Moreover, other innovative and analytical approaches may further elucidate users’ perceptions and experiences, such as text mining approaches in product development employed by Lin (2018). Last, a future study may also look into the effect of other demographic factors (e.g., ethnicity) on weather perception among respondents, as KSA hosts migrants from all over the world.
6. Conclusions
This study represents a step toward more data-based evidence about the needs, opinions, and usage of weather forecast services in KSA by utilizing the survey research approach. On the basis of 315 responses, this study reports on the perception of the citizens of KSA with regard to the importance of the weather forecast information as well as their opinion of the importance of a set of weather forecast features and information and their beliefs with regard to the availability of these features and information in the current weather apps. In general, it seems that KSA citizens take the weather forecast information seriously and adjust aspects of their day-to-day life based on these forecasts. People tend to dress and change their outdoor activity plans on the basis of the weather forecast information.
KSA citizens indicated that most surveyed weather forecast services and features—such as having detailed forecast information, weather alerts, and the ability to visualize information on maps and pinpoint exact locations—as important and necessary. The results from the survey indicated a mix of opinions with regard to the belief of the existence of such weather forecast information and features in the currently available weather apps; a large portion of the participants, 32%–47%, stated that they have no information about the existence of these services and features. The findings in this study will serve as a base for more detailed studies that consider the opinion and perspective of a larger and more diverse population of the citizens of KSA and will try to understand the needs in more detail through other research methods such as interviews and focus groups.
Acknowledgments
The authors thank the Deanship of Scientific Research at Jazan University for financially supporting this work (Project W41-039). There is no conflict of interest.
Data availability statement
The authors confirm that the data that support the main findings of this study are available within the document. The full raw data can be accessed upon request from the corresponding author (Abdulrahman Khamaj).
REFERENCES
Al Arabia, 2016: Saudi Arabia’s vision 2030. Al Arabiya News, https://english.alarabiya.net/en/perspective/features/2016/04/26/Full-text-of-Saudi-Arabia-s-Vision-2030.html.
Argyle, E., J. Gourley, Z. Flamig, T. Hansen, and K. Manross, 2017: Toward a user-centered design of a weather forecasting decision-support tool. Bull. Amer. Meteor. Soc., 98, 373–382, https://doi.org/10.1175/BAMS-D-16-0031.1.
Ash, D., L. Ronald, and C. Gregg, 2014: Tornado warning trade-offs: Evaluating choices for visually communicating risk. Wea. Climate Soc., 6, 104–118, https://doi.org/10.1175/WCAS-D-13-00021.1.
Bean, H., J. Sutton, B. F. Liu, S. Madden, M. M. Wood, and D. S. Mileti, 2015: The study of mobile public warning messages: A research review and agenda. Rev. Commun., 15, 60–80, https://doi.org/10.1080/15358593.2015.1014402.
Bryant, M., W. Smart, and S. Wilde, 2015: Consumer influences into weather apps usage on smartphones: Key highlights. Proc. Advances in Business-Related Scientific Research, Milan, Italy, GEA College, 8–15.
Chan-Olmsted, S., H. Rim, and A. Zerba, 2013: Mobile news adoption among young adults: Examining the roles of perceptions, news consumption, and media usage. J. Mass Commun. Quart., 90, 126–147, https://doi.org/10.1177/1077699012468742.
Cools, M., and L. Creemers, 2013: The dual role of weather forecasts on changes in activity-travel behavior. J. Transp. Geogr., 28, 167–175, https://doi.org/10.1016/j.jtrangeo.2012.11.002.
Darfaoui, E. M., and A. A. Assiri, 2010: Response to climate change in the Kingdom of Saudi Arabia. Director General of the Department of Natural Resources Rep., 17 pp., http://www.fao.org/forestry/29157-0d03d7abbb7f341972e8c6ebd2b25a181.pdf.
Demuth, J. L., J. K. Lazo, and R. E. Morss, 2011: Exploring variations in people’s sources, uses, and perceptions of weather forecasts. Wea. Climate Soc., 3, 177–192, https://doi.org/10.1175/2011WCAS1061.1.
Drogalis, T., E. Keyes, and L. Dhyani, 2015: The Weather Channel application usability test. De Paul University Rep., 13 pp., http://elkeyes.com/wp-content/uploads/2015/05/WeatherChannelUTResults.pdf.
Drost, R., 2013: Memory and decision making: Determining action when the sirens sound. Wea. Climate Soc., 5, 43–54, https://doi.org/10.1175/WCAS-D-11-00042.1.
FloodList, 2019: Saudi Arabia–Deadly flash floods in Eastern Province after 43mm of rain in 30 minutes. FloodList News, 29 October, http://floodlist.com/asia/saudi-arabia-floods-hafralbatin-october-2019.
Gökçearslan, Ş., F. K. Mumcu, T. Haşlaman, and Y. D. Çevik, 2016: Modelling smartphone addiction: The role of smartphone usage, self-regulation, general self-efficacy and cyberloafing in university students. Comput. Hum. Behav., 63, 639–649, https://doi.org/10.1016/j.chb.2016.05.091.
Gul, F. A., and C. M. Ali, 2016: Saudi crane collapse Masjid al-Haram (lack of safety). J. Soc. Sci. Humanit. Res., 1, 129–140.
Hayden, M. H., S. Drobot, S. Radil, C. Benight, E. C. Gruntfest, and L. R. Barnes, 2007: Information sources for flash flood warnings. Environ. Hazards, 7, 211–219, https://doi.org/10.1016/j.envhaz.2007.07.001.
Heath, R. L., J. Lee, M. J. Palenchar, and L. L. Lemon, 2018: Risk communication emergency response preparedness: Contextual assessment of the protective action decision model. Risk Anal., 38, 333–344, https://doi.org/10.1111/risa.12845.
Jon, I., H. Shihkai, and L. Michael, 2018: Perceptions and reactions to tornado warning polygons: Would a gradient polygon be useful? Int. J. Disaster Risk Reduct., 30, 132–144, https://doi.org/10.1016/j.ijdrr.2018.01.035.
Khamaj, A., and Z. Kang, 2018: Usability evaluation of mobile weather hazard alert applications. Ind. Syst. Eng. Rev., 6, 21–40, https://doi.org/10.37266/ISER.2018v6i1.pp21-40.
Khamaj, A., Z. Kang, and E. Argyle, 2019: Users’ perceptions of smartphone weather applications’ usability. Proc. 63rd Human Factors and Ergonomics Society Annual Meeting, Los Angeles, CA, Human Factors and Ergonomics Society, 2216–2220, https://doi.org/10.1177/1071181319631098.
Lazo, J. K., R. E. Morss, and J. L. Demuth, 2009: 300 billion served: Sources, perceptions, uses, and values of weather forecasts. Bull. Amer. Meteor. Soc., 90, 785–798, https://doi.org/10.1175/2008BAMS2604.1.
Lin, K. Y., 2018: A text mining approach to capture user experience for new product development. Int. J. Ind. Eng., 25, 108–121.
Lindell, M. K., and R. W. Perry, 2012: The protective action decision model: Theoretical modifications and additional evidence. Risk Anal., 32, 616–632, https://doi.org/10.1111/j.1539-6924.2011.01647.x.
Na, K., and G. M. Lee, 2016: Agent-based simulation of emergency evacuation for nuclear plant disaster. Int. J. Ind. Eng., 23, 445–448.
Nayebi, F., J. M. Desharnais, and A. Abran, 2012: The state of the art of mobile application usability evaluation. Proc. 25th Annual IEEE Canadian Conf. on Electrical and Computer Engineering, Montreal, QC, Canada, IEEE, 1–4, https://doi.org/10.1109/CCECE.2012.6334930.
Pappenberger, F., H. L. Cloke, D. J. Parker, F. Wetterhall, D. S. Richardson, and J. Thielen, 2015: The monetary benefit of early flood warnings in Europe. Environ. Sci. Policy, 51, 278–291, https://doi.org/10.1016/j.envsci.2015.04.016.
Phan, M. D., B. E. Montz, S. Curtis, and T. M. Rickenbach, 2018: Weather on the go: An assessment of smartphone mobile weather application use among college students. Bull. Amer. Meteor. Soc., 99, 2245–2257, https://doi.org/10.1175/BAMS-D-18-0020.1.
Purcell, K., 2011: Half of adult cell phone owners have apps on their phones. Pew Internet & American Life Project Rep., https://www.pewresearch.org/internet/2011/11/02/half-of-adult-cell-phone-owners-have-apps-on-their-phones/.
Rianti, E., S. D. Rizky, and F. H. Nugraha, 2020: Evaluation of the satisfaction of users of weather forecast systems with the service quality method. Sinkron, 4, 130–140, https://doi.org/10.33395/sinkron.v4i2.10512.
Sensor Tower, 2020: Weather—The Weather Channel. https://sensortower.com/ios/US/the-weather-channel-interactive/app/weather-the-weather-channel/295646461/overview.
Silver, A., 2015: Watch or warning? Perceptions, preferences, and usage of forecast information by members of the Canadian public. Meteor. Appl., 22, 248–255, https://doi.org/10.1002/met.1452.
Silver, A., and C. Conrad, 2010: Public perception of and response to severe weather warnings. Meteor. Appl., 17, 173–179, https://doi.org/10.1002/met.198.
Statista, 2019a: Number of smartphone users in Saudi Arabia from 2017 to 2023 (in millions). https://www.statista.com/statistics/494616/smartphone-users-in-saudi-arabia/.
Statista, 2019b: Market reach of the most popular mobile app categories in the United States as of September 2019. https://www.statista.com/statistics/579302/top-app-categories-usa-reach/.
Sutton, J., and L. M. Fischer, 2021: Understanding visual risk communication messages: An analysis of visual attention allocation and think aloud responses to tornado graphics. Wea. Climate Soc., 13, 173–188, https://doi.org/10.1175/WCAS-D-20-0042.1.
Tavakol, M., and R. Dennick, 2011: Making sense of Cronbach’s alpha. Int. J. Med. Educ., 2, 53–55, https://doi.org/10.5116/ijme.4dfb.8dfd.
Tsai, Y. F., J. L. Hsieh, W. G. Teng, T. W. Hou, C. P. Freg, and Y. C. Tsao, 2018: Discovering disaster events from social media streams. Int. J. Ind. Eng., 25, 685–701.
Weiss, A., and J. G. Robb, 1986: Results and interpretations from a survey on agriculturally related weather information. Bull. Amer. Meteor. Soc., 67, 10–15, https://doi.org/10.1175/1520-0477(1986)067<0010:RAIFAS>2.0.CO;2.
Wong, T. F., and Y. Y. Yan, 2002: Perceptions of severe weather warnings. Meteor. Appl., 9, 377–382, https://doi.org/10.1017/S1350482702003110.
Wu, H.-C., M. K. Lindell, and C. S. Prater, 2015: Process tracing analysis of hurricane information displays. Risk Anal., 35, 2202–2220, https://doi.org/10.1111/risa.12423.
Wu, H., S. Arlikatti, A. Prelog, and C. Wukich, 2017: Household response to flash flooding in the United States and India: A comparative study of the 2013 Colorado and Uttarakhand disasters. Understanding Vulnerability, Building Resilience: Responses to Disasters and Climate Change, M. Companion and M. Chaiken, Eds., CRC Press, 37–48.
Youssef, A. M., S. A. Sefry, B. Pradhan, and E. A. Alfadail, 2016: Analysis on causes of flash flood in Jeddah city (Kingdom of Saudi Arabia) of 2009 and 2011 using multi-sensor remote sensing data and GIS. Geomatics Nat. Hazards Risk, 7, 1018–1042, https://doi.org/10.1080/19475705.2015.1012750.
Zabini, F., 2016: Mobile weather apps or the illusion of certainty. Meteor. Appl., 23, 663–670, https://doi.org/10.1002/met.1589.
