Introduction

Developments in information and communication technologies (ICTs) offer, at first glance, real chances for the emergence and operation of a modern, digital agora. In ancient times, the agora was an open public assembly place where people gathered to share and discuss ideas and concerns related to politics, everyday affairs, or other issues. This exchange of ideas also led to social changes; not surprisingly is the agora of Athens called the cradle of democracy (Mark 2021). New electronic public spaces enabled by ICTs can mimic certain functions of the ancient agora by (1) enabling citizens’ direct participation in social processes through virtual spaces; (2) allowing citizen involvement in decision-making; and (3) furthering dialog among citizens and also between citizens and policymakers (Kamps 2000). Therefore, these spaces promise participation and collaboration opportunities for society, as already indicated to some extent in fields such as politics (Kirk and Schill 2011; Parycek et al. 2017), religion (Galang and Macaraan 2021) and science (Olszewski and Wendland 2021). In terms of science, the digital agora could emerge as a place where scientific knowledge is generated, shared, discussed and acted upon through public participation, thus becoming an auspicious venue for citizen science.

Citizen science (CS)

While CS is difficult to describe with one single definition, at its core, it refers to some form of public participation in scientific research (Haklay et al. 2021). The degree of public participation can vary, usually classified into four approaches depending on the level of public involvement in the scientific process: a) contributory approach, when citizen scientists contribute only to data collection; b) collaborative approach, when citizen scientists are also involved in data interpretation; c) co-created approach, when citizen scientists are involved in the entire scientific process, including research structure or d) citizen-led approach – when citizen scientists take up a leading role throughout the whole scientific process (Marks et al. 2022). The levels proposed by Haklay (2013) can be identified in this scale: Crowdsourcing, Distributed Intelligence, Participatory Science and Extreme Citizen Science. The main difference is given by the activities that volunteers can conduct within the last level of engagement: initiation and coordination of a CS project, and by the dissemination of results.

Throughout history, there have been several instances where citizens volunteered to engage in different stages of scientific processes to further our understanding of our environment, whether related to meteorology, astronomy, or oceanology (García et al. 2021). The term CS, as we came to know it today, was first recorded in 1989 to describe volunteers’ engagement in scientific data generation related to rain acidity levels in the USA, an issue of political relevance at that time (Haklay et al. 2021). However, the benefits of CS continue beyond the resources that citizens can contribute to science and its contribution to community capacity building for decision-making and influencing policies (Van Vliet and Moore 2016). It can also serve as a vehicle for science communication and education, offering a better understanding for citizen scientists of science and the academic research process (Resnik et al. 2015). Contributing to tasks set out by CS projects is merely one form of learning. Participants can also enhance their understanding of science and specific topics by using external resources and project documentation, sharing their creations, or interacting with others (Jennett et al. 2016), thus improving their scientific literacy. To what extent can CS improve scientific literacy is difficult to measure (Cronje et al. 2011). Nonetheless, several studies report on such improvements (Brovelli et al. 2020; Haklay, 2021; Kloetzer et al. 2021; Robinson et al. 2018). Social participation, defined as active engagement in discussions between project participants, proved to be an important variable in promoting scientific literacy through CS (Price and Lee 2013). Therefore, making use of ICTs to create venues that facilitate such social interactions is worth considering.

Citizen science mobile apps (CS apps)

ICTs have enabled new forms of engagement for scientists and citizens through virtual communities of practice (Price and Lee 2013; Skarlatidou et al. 2019) paving the path for digital CS. In the recent decade, many online CS projects emerged, partly due to wider internet access and smartphones that are increasingly performant and affordable (Kloetzer et al. 2021). Thus, mobile apps can achieve diverse participation in CS (Paleco et al. 2021) as well as offer simplified and varied data collection and networking opportunities (Andrachuk et al. 2019; Ferster and Coops 2013). Geographic Information Systems (GIS) add to the contributions of smartphones and their apps. Without the opportunities presented by these, the perspectives of CS would be far less optimistic and promising. Among these, location tracking is perhaps the most significant. Any CS component concept (e.g. Participation GIS, Community Based Monitoring, Participatory Research, Community Engagement, and Volunteered Geographic Information) uses the advantages offered by the GIS. This is the argument for which the inclusion of Volunteered Geographic Information (VGI) as a core concept in CS is not without justification, as it serves as a fundamental principle underlying numerous activities that fall within this category. Moreover, the concept of “geographic citizen science” emerges at the intersection of CS and VGI, a concept defined as “scientific work undertaken by members of the general public where the generated data have a deliberate and explicit geographic aspect” (Haklay 2021). Due to its advantages, VGI is widely exploited by researchers as a decision-making support tool in various fields, being easily collected and includes a large volume of spatial-temporal data as compared to the traditional sources (Howe 2021; Yan et al. 2020). In addition, the platform format makes possible geoparticipation, which is similar to the real participation of the residents in decision-making, contributing to a digital agora (Olszewski and Wendland 2021). Among the principles that can make it operational, we mention: “thinking through people” (Howe 2021) based on “citizens as sensors” (Goodchild 2007) and “no one left behind” (Skarlatidou and Haklay 2021).

Scientific interest towards CS apps

CS is a topic of interest for the scientific community in many fields: Ecology (20.9%), Environmental Sciences (20.6%), Biodiversity Conservation (14.2%) Zoology, Public Environmental & Occupational Health, Water Resources, Computer - Science Information System, Geosciences - Multidisciplinary, Education & Educational Research, Green & Sustainable Science & Technology, Astronomy & Astrophysics, Geography Physical, Oceanography, Meteorology & Atmospheric Sciences, etc. The number of research papers in the Web of Science (WoS) database containing “Citizen Science” in their topic increased starting with the 2000s. The year 2022 recorded the peak (n = 1341) within the time frame examined (Fig. 1). The year 2010 is of particular interest from the scientometric point of view. Starting this year, the concept was no longer covered by the Citizen Science Paradox – a phenomenon referring to the lack of the “Citizen Science” concept in papers published prior to 2010, before the CS concept formalization from 2009 facilitated its referencing (Bergerot 2022). Therefore, prior to 2010, the search results in WoS listed a much lower number of publications on CS. A similar ascending trend, but with significantly higher values (+33 – +212%), is noticed for research papers containing the topic “mobile phone applications”. Starting with 2007, the topic “mobile phone apps” became more featured in literature, while the peak in the number of studies was reached in 2021 (n = 1994). The similar evolution of the number of studies addressing the two concepts indicates a promising prerequisite for CS mobile phone apps, with smartphones becoming more efficient and affordable, which can enable the diversity and accessibility of CS apps.

Fig. 1: Scientific interest for CS and mobile apps throughout the years.
figure 1

The evolution of the number of studies on the topics of “citizen science” and “mobile phone app”, according to WoS (17.08.2023).

Full size image

The synthesis of progress regarding the two concepts becomes a topic of literature reviews starting with 2015 for CS and 2013 for mobile apps. The highest values are recorded during these years, while the differences follow the same trend (+85–+176%) as in the case of the analyzed studies. Of these, 26 address the use of mobile phone apps in CS (Table 1). These studies assess the phenomenon either at global level (Encarnação et al. 2021; Howard et al. 2022; Johnson et al. 2020) or at regional level (Kipp et al. 2019; Wakeling et al. 2022), but none of them considers the entire range of domains where CS and mobile phone apps are approached together. The most comprehensive regard the implications in assessing the environmental factors (Andrachuk et al. 2019) and the One Health Spectrum (McNeil et al. 2022).

Table 1 Use of CS apps in various fields.
Full size table

Therefore, in this research, we aim to describe the landscape of CS apps. To this end, we pursue two specific research questions:

RQ1: What is the current landscape of using mobile apps in CS?

RQ2: What are the possible future trends in this field?

By the current landscape of CS apps, we understand first the range of applicability (the number of domains and subdomains that attract the most interest) and second, the associated content categories. The second research question builds on these two and extends the possibility of their future application.

Methodology

To address the two research questions, we carried out an extensive bibliographic search and literature review. In what follows we shall outline a detailed description of the steps involved in identifying and processing the bibliographic materials.

Literature reviews have become increasingly valuable in recent years, as platforms such as WoS Core Collection and Scopus (among others) continue to expand their titles and offer more robust search capabilities. However, this does not mean that limitations have been completely eliminated, particularly when speaking about concepts such as citizen science. Some notable limitations stand out, particularly those that influence the results of the most common form of search query - the topic (title, abstract, author keywords, and keywords plus). We are talking about the false positive and false negative results, whose occurrence conditions were described by Kullenberg and Kasperowski (2016).

The false positive results refer to the generation of some titles with no connection to the subject of the search. The most common causes are the multiple significances of some terms (e.g. acronyms) and the presence of WoS Keyword Plus in the topic (Garfield and Sher 1993). The false negative results refer to the absence of some relevant titles in the results of the search, which may be caused by the omission of certain search terms (e.g. out of ignorance) or by the Citizen Science Paradox (Bergerot 2022). One way to address the first issue is by manually removing false positive results. The process may entail initially analyzing the topic elements and subsequently examining the full-text documents, with the condition of not having a large number of titles. This action often suffices to effectively eliminate blatant/flagrant false positive results. The second issue - the numerical reduction of the false positive results - is a more challenging problem to address. This does not mean that we should abandon the use of scientific databases queries, but rather optimize our working methods and broaden search capabilities within these databases, such as identifying synonyms of the searched terms.

Literature analyses on topics related to CS can be useful if we focus on WoS titles published after 2010 (the titles published before 2010 fall into the Citizen Science Paradox). We consider that it would be appropriate to integrate the recursive searches Kullenberg and Kasperowski (2016) in the workflow summarized in the PRISMA flow chart (Page et al. 2021). Recursive searches (RS) can help identify terms commonly associated with a particular concept, which can increase the percentage of representative titles generated and ultimately reduce the number of false positive results.

This method is particularly useful for our research for two reasons: a) the relationship between CS and mobile apps is a relatively new area of research in the international scientific community, and b) the concepts related to CS and mobile apps are diverse and not evenly distributed across the existing literature reviews on the topic (Bautista-Puig et al. 2019; Bergerot 2022; Follett and Strezov 2015; Kullenberg and Kasperowski 2016). These factors make it more challenging to identify relevant titles over a certain period, even when focusing only on recent publications (in our case 2017–2022).

Recursive searches have enabled us to highlight the concepts more frequently associated with citizen science. When inserted in the search expression, these have increased the chances of identifying relevant titles. The work phases prior to the systematic literature analysis are presented in Fig. 2 PRISMA 2020 flow chart, adapted after Page et al. (2021).

Fig. 2: The work phases of the literature review.
figure 2

The PRISMA 2020 flow chart outlines the methodological steps of the CS apps literature review.

Full size image

The methodological steps are the following:

  1. a.

    The initial search was conducted in WoS Core Collection, on 4 December 2022, by the expression “Citizen Science”, targeting all studies in English.

  2. b.

    The resulting titles (n = 6259) were subject to a bibliometrical analysis (Donthu et al. 2021), co-occurrence (full counting), in VOSviewer 1.6.18, based on author keywords (Kullenberg and Kasperowski 2016). From the keywords occurring at least 5 times, some were selected in order to compose the expression that would lead to the second search. The terms selected refer to most of the CS activities, identified by Haklay et al. (2021) and they follow approximately the criteria proposed by Kullenberg and Kasperowski (2016) (Supplementary Table S1).

  3. c.

    The second search was conducted in WoS Core Collection, on 26 December 2022, by an expression composed based on the key words identified in stage b). The following were used: “” (the expression was assessed as individual word) and * (words/expressions that contained the indicated root were generated as results) to offer a wider range to the search expression. Among the additional criteria we mention the publication language (English) and document type (Article, Review Article and Early Access) (Kraus et al. 2022).

  4. d.

    The additional criteria mentioned in step c) have eliminated only partially the unwanted titles. Therefore, it was necessary to manually select from the generated titles (n = 636) the Book chapter (n = 12) and Proceedings paper (n = 109) documents. The titles in the peer review with no JCRs’ 5-Year impact publications (n = 44) were eliminated (Loureiro et al. 2020), those before 2018 (except the review articles) (n = 113), doubles (n = 3) and titles irrelevant for our study (n = 16). A small number (n = 3) of titles could not be obtained in full text format. After reading the titles and abstracts, (n = 335) titles were identified. Most were research articles (n = 312), while the rest (n = 23) were review articles (Supplementary Table S1).

  5. e.

    The first analysis of the entire text had two objectives: i) identify the extent to which the objectives of the review type studies coincide with those of our study, wishing to avoid potential double cases; and ii) outline the major categories that the other articles could fit in. At this stage, we identified the review type studies present in the database as research article (n = 9). Therefore, the final number of studies in this category reached (n = 31).

  6. f.

    303 studies were subject to systematic literature analysis (for the detailed list of the articles see Supplementary Table S1).

  7. g.

    Based on these studies we outlined the current landscape of using mobile apps in CS and identified future research perspectives.

Results and interpretation

The application of CS in various domains

As Fig. 3 shows, CS apps cover a great scope. Some are designed to observe the surrounding environment, either by monitoring specific ecological components or by keeping track of environmental risks and pollution. But there are also other domains of application, such as improving well-being or addressing challenges of CS.

Fig. 3: The current landscape of CS apps.
figure 3

The Sankey diagram depicts the main domains and subdomains of CS applicability and their associated content categories.

Full size image

Monitoring environmental risk and public health (n  =  151)

Engaging citizens in environmental monitoring is the most common application of CS tools. Water quality indicators are the most common environmental factors monitored through CS apps when it comes to environmental pollution (Arienzo et al. 2021; King et al. 2022; Tedla et al. 2022; Ward-Paige et al. 2022; Zheng et al. 2022). There are also other applications, such as measuring radiation (Bottollier-Depois et al. 2019) or noise levels (Zipf et al. 2020), shoreline change monitoring (Jaud et al. 2020) or waste monitoring (Córdova et al. 2022; Scardino et al. 2022). Crowdsourced data on the quality of various environmental factors and populations’ health can help scientists and policymakers understand environmental risks better, thus proposing more adequate solutions for improving community resilience. For example, in Bangalore, India, the PPAC Waste Tracker was used by citizen scientists to collect data on issues related to the city’s waste management. Using visual mapping techniques, the data was presented to local officials and service providers (Nagendra et al. 2019). CS apps can also raise awareness of environmental problems by engaging citizens in observing and documenting issues. To improve citizen engagement in data collection, some apps use gamification techniques. For example, through Globe Observer, different data collection challenges are set up for citizen scientists to participate in (Amos et al. 2020). One of these challenges, the Cloud Spring Data Challenge, managed to collect 55,639 cloud observations in just a month (Robles et al. 2020).

Citizen scientists can also contribute data on hazards, like landslides, earthquakes, wildfires, or floods, fulfilling needs from different phases of the disaster management cycle. CS apps can enable collaborative risk assessment through initiatives like participatory landslide inventory mapping (Anshori et al. 2022; Kocaman and Gokceoglu, 2019; Sellers et al. 2021). Also, citizens can participate in hazard and risk monitoring to improve early warning (Allen et al. 2020; Ferri et al. 2020; Finazzi 2020b; Kong et al. 2019) or situational awareness (Saran et al. 2020; Yalcin et al. 2020). Some apps are designed to raise public awareness. In Poland, for example, citizens can record digital floodmarks as substitutes for historical landmark plaques to keep the memory of extreme events alive (Szczepanek et al. 2022). Risk prevention and preparedness behaviors can also be encouraged through CS apps. Such is the case of Smoke Sense, an app designed to engage individuals in recording environmental changes and their health when exposed to wildfire smoke. The app also contains practical information on health protection, and the data recorded by citizen scientists is used to improve health risk communication (Rappold et al. 2019).

CS apps in the public health domain are mainly designed to monitor epidemics and zoonotic or plant allergens (Low et al. 2021; Nadini et al. 2020; Silver et al. 2020b, 2020a). Another application of such apps is testing methods and techniques that allow the recognition of disease vectors (Cochero et al. 2022).

Monitoring biodiversity and reconnecting people with nature (n  =  75)

Another common application of CS apps is encouraging citizens to observe nature and collect data on biodiversity. Many terrestrial and marine species are currently monitored by citizen scientists worldwide, such as birds (Brooks et al. 2022; Wessels et al. 2019), mosquitoes (Braz Sousa et al. 2020; Carney et al. 2022; Eritja et al. 2019; Low et al. 2021), insects (Chen et al. 2020; Cochero et al. 2022), jellyfish (Bellido et al. 2020), plants (Boho et al. 2020) or forest vegetation (Danqi et al. 2021; Kowler et al. 2020; Tatsumi et al. 2022).

The scope of biodiversity monitoring apps is diverse. For example, some CS apps are designed to collect data on local and invasive species, aiding scientists and decision-makers in developing adequate environmental policies to conserve local flora and fauna (Crocker et al. 2020; Kays et al. 2022; Marambio et al. 2021; Stenhouse et al. 2020) or manage invasive species (Benjamins et al. 2018; Geller et al. 2021; Malek et al. 2019). Apps also often focus on educating users about terrestrial and marine flora and fauna, thus raising awareness of the importance of conserving biodiversity. For example, in Australia, citizen scientists who used the TurtleSAT app reportedly adopted different turtle-friendly actions, indicating behavior and attitude changes once they became more knowledgeable of turtle species decline (Santori et al. 2021). Some apps also increase their users’ knowledge on different plant or animal species by automatically identifying plant species using AI-based recognition systems (Boho et al. 2020; Kress et al. 2018; Mäder et al. 2021; Patel et al. 2020). Public interest towards apps that utilize such recognition systems is seemingly high. For example, FungiVision, an AI-based fungi species recognition system, increased fourfold the number of citizen scientist engaged in data collection in Denmark (Picek et al. 2022). CS apps aimed at environmental education can be designed also to specifically improve classroom experience and, in many instances, such apps go beyond enhancing scientific literacy to raising environmental citizens (Echeverria et al. 2021; Klütsch et al. 2021; Young et al. 2021).

Improving well-being (n  =  58)

Citizen scientists can also contribute to making the built environments more sustainable by testing and using smart monitoring solutions. They can track, for example, thermal comfort indicators in intelligent buildings (Chiesa et al. 2022) or can contribute data on smart cities by identifying elements that impact their well-being (Chrisinger and King 2018). Crowd-sensed and crowd-sourced data can also be used for mapping urban environments by creating accurate surface maps (Fan et al. 2019), 3D city models (Bshouty et al. 2019), or by exploring heat patterns (Yin et al. 2021). Such information can help decision-makers to find sustainable solutions for land administration (Du et al. 2021; Olteanu-Raimond et al. 2018), urban green space planning and management (Heikinheimo et al. 2020; Schrammeijer et al. 2022), human mobility improvement (Hologa and Riach 2020; Howe 2021), or heritage building management (Brigham et al. 2018; Lamas et al. 2021).

CS apps can also engage users in a healthier and more active lifestyle, by encouraging CS platform users to gather data while performing pastime activities such as tracking, recreational fishing, or active walking. This crowed-sourced data can improve scientists understanding of outdoor recreational activities (Muñoz et al. 2019; Tuckett et al. 2018) or recreational fishery management (Gundelund et al. 2020, 2022; Nieman et al. 2020).

Some domains of applicability are less well represented:

Primary and secondary economy (n  = 14)

Is such a domain focusing mainly on different aspects of agricultural, such as food security (Hampf et al. 2021), as well as understanding how to use new technologies to further sustainable agriculture (Laso Bayas et al. 2020; Posadas et al. 2021; Thompson et al. 2019). Other subdomains of economy are fishery management (Jeffers et al. 2019; Silva et al. 2022) and resource management (Bernedo Del Carpio et al. 2021; Maynard et al. 2020; Ogbe and Lujala, 2021).

Limitations and barriers (n  =  3)

To CS apps is another domain scarcely studied at present. However, it shows a real potential for development in the future by addressing topics like data safety (Tovino 2020b, 2020a) or cartographic vandalism on user-generated data in the era of location-based games (Juhász et al. 2020).

Linguistics (n  =  2)

Is another domain of applicability, that enables recording free speech or reconstructing linguistic landscapes (Hilton 2021; Purschke 2021).

The application of CS ranges also in its spatial scale from regional to global applicability. The majority operates on the national level, encouraging citizen data contribution to observe and analyze environmental issues on country-level, for example, the SnailSnap app, designed to monitor urban heat islands in the Netherlands by collecting pictures of the land snails (Kerstes et al. 2019). National application of CS can also support tracking and sharing information on the national biodiversity, as in the case of FrogID, meant to document frog diversity in Australia’s urban regions (Callaghan et al. 2021; Rowley et al. 2019; Rowley and Callaghan, 2020). Another application is offering nationwide services to improve safe living, such as the Earthquake Network app from Chile (Finazzi 2020a). CS can also be applied in smaller regions within a country to raise awareness of local environmental issues. For example, BAYSICS is designed to promote environmental learning in the context of climate change in Bavaria, the south-eastern part of Germany (Batsaikhan et al. 2020). Small-scale application of CS can also improve spatial planning by gathering insight into how residents perceive their immediate surroundings, such as urban green spaces (Heikinheimo et al. 2020; Schrammeijer et al. 2022). Application of CS can also cross borders to study specific regions, as does the Invasive Alien Species in Europe mobile phone app. Besides tracking invasive land species in Europe (Schade et al. 2019) and invasive freshwater species in the Lower Danube Region (Trichkova et al. 2021) these apps also set out to make data on the issue accessible to the relevant policymaking at EU level. There are also global CS initiatives, such as the eOcean or iNaturist platforms. These platforms connect scientists and citizen scientists worldwide, allowing them to initiate new monitoring projects or participate in already-established ones.

Content categories of CS apps

The referenced literature reveals three major content categories (Fig. 3), each of them with a certain applicability, such as: (1) use of specific monitoring tools, (2) techniques and methods associated with CS mobile apps; (3) participatory engagement of citizens by means of mobile apps.

Use of specific monitoring tools (n = 141)

A series of tools, collaborative platforms and mobile apps exist that facilitate the collection of data by citizens. One of the most frequently used platform worldwide is OpenStreetMap, due to its open-source nature (Cao et al. 2020; Dumedah et al. 2022). With the rise of CS, other popular platforms emerged such as Mosquito Alert, GlobeObserver, iNaturalist, FrogID, which engage millions of citizens in environmental observations (Altrudi 2021; Amos et al. 2020; Callaghan et al. 2021; Carney et al. 2022; Chen et al. 2020; Eritja et al. 2019; Low et al. 2021). Thus, citizen scientists can collect data, samples and observations with geolocation and photos (Arienzo et al. 2021; Cakmak et al. 2021). The proliferation of data acquisition sources for scientific analyses, especially from media or internet sources has also led to the need of introducing some more rigorous selection criteria (Bellido et al. 2020). Beyond data collection, associated metadata also contribute with additional information (Boho et al. 2020). The success of using some CS tools such as Picture Pile, derives also from meeting national standards for statistical data quality assurance, an important step in adopting the FAIR data principles (Findable, Accessible, Interoperable and Reusable). The mentioned app includes several data categories such as: satellite images, ortophotoplans, photos with geolocation elements to measure various sustainable development indicators (Fraisl et al. 2022).

Techniques and methods associated with CS mobile apps (n  = 114)

Using data collected from untrained individuals in prognosis methods can contribute to reducing the negative effects of certain risk phenomena by improving prevention systems (Annis and Nardi 2019; Bogdos and Manolakos 2019; Finazzi 2020a). The great drawback of VGI derives from the fact that this is considered qualitative information. Therefore, it requires methods for data filtering and data accuracy testing, in some cases AI models (Annis and Nardi, 2019; Brigham et al. 2018; Castro-Gutiérrez et al. 2022; Omodior et al. 2021; Pepe et al. 2021). It also needs methods for testing the accuracy of mobile devices (Blair et al. 2018; Cabrera et al. 2021).

As a result, some authors have suggested advanced techniques of automated observation modeling that eliminate the requirement for manual validation by humans (Botella et al. 2018). Others have proposed the development of platforms of varying complexity integrated with IoT (Internet of Things) sensors, to create networks capable of collecting larger data sets and calculating specific indexes (Adhane et al. 2021; Caputo et al. 2020; Chiesa et al. 2022; Jackowiak et al. 2022; Kays et al. 2022; Polineni et al. 2022). New apps for data collection have been developed within other studies, such as: Nutrient App for monitoring water quality parameters (Costa et al. 2020), OpenRadiation for measuring radioactivity (Bottollier-Depois et al. 2019), ClimApp for personalized thermal models (Kingma et al. 2021), NoiseExplorer for noise measurement (Garg et al. 2020), StopTheSpread for flu prevention (Nadini et al. 2020), CREDO Detector app for registering different forms of radiation (Bibrzycki et al. 2020). The development of research infrastructure elements to support the monitoring of data is also mentioned in some studies, either when speaking of mobile monitors (Chen et al. 2020) or when speaking about scanners (Tatsumi et al. 2022).

The literature analysis also highlighted some studies by which specific working methods have been created to be used in monitoring. The photo-identification methods serve the identification of specific species or phenomena based on the characteristics found in photographic content. To give a few examples, we mention individual recognition through photo-identification to monitor flapper skates (Benjamins et al. 2018), artificial light classification through color analysis from photos (Muñoz-Gil et al. 2022). Developing specific algorithms to analyze CS data enables further applications, such as the simulation of colorimetric analyses to monitor water quality (Li et al. 2022) or building height calculation to attribute data for 3D building models (Bshouty et al. 2019). Executing other models proposed by various authors also helps to automatically classify data by providing feedback to the app users (Pataki et al. 2021).

Participatory engagement of citizens by means of mobile apps (n  =  48)

Mobile apps are used as work tools for collecting spatial and temporal data and achieving information from citizens. We take for example Our Voice app, used to document the citizen scientists’ experiences regarding the lifestyle of the respective community, a materialization of the contributory approach (Afaneh et al. 2022). Furthermore, this tool found also in other studies (González et al. 2022; Odunitan-Wayas et al. 2020; Tuckett et al. 2018) combines the contributory approach with the mobile technology and collection methods of the data provided by citizens (audio recordings, images). Such an effort results in multidisciplinary outputs. The input provided by researchers resulted in determining the factors and experiences of the communities affected by hurricane Katrina, while the interview participants have offered information on the effects of Katrina felt by people, from material losses to the trauma they have been through (Afaneh et al. 2022). Another advantage of using this app comes from community-based empowerment. Citizens become thus agents of change in their own communities (Coudel et al. 2022; Tuckett et al. 2018).

Data collection based on the collaborative model, associated with the visualization of data may be exemplified by the NoiseSpy app, for measuring the comfort degree inside some spaces (Rozzi et al. 2022). In a co-created model, considering the living environment within a community, another study focused on the elderly citizens group. By means of an app, the seniors have recorded and documented photographically, for 3 months, the green spaces from their community (Barrie et al. 2019). In other situations, some agencies have chosen the creation of crowdsourced data collection platforms giving up on the third-party data, as the case of Cycletracks app, which collects VGI data (Alattar et al. 2022; Beza et al. 2018).

Another challenge of CS comes from engaging the participants and maintaining citizen empowerment. In the case of some apps monitoring environmental aspects, the volunteers have already proven an increasing a priori interest for environmental issues (Afaneh et al. 2022). Among the advantages of using citizen science, we mention community-based monitoring, where citizens may take the role of scientists (David Izquierdo-Gómez 2022; Fraisl et al. 2022).

The educational environment (pre-university, university) is responsible for the building of individual and collective behavior of pupils and students. From this reason, CS may contribute to the building of a supportive behavior, by engaging young people in monitoring various aspects of society (Daum et al. 2022; Echeverria et al. 2021; Hammami and Katapally 2022; Katapally 2022). The advantages of these involvements translate into: 1) becoming familiar with new technologies for data collection, 2) contribution of the data and proposed solutions, 3) visualization of data collected by youth by stakeholders such as authorities (Odunitan-Wayas et al. 2020).

Participation of volunteers in interviews and surveys is conditioned by the mechanisms stimulating their involvement. From this point of view, there are two possibilities: self-reported motivations, such as curiosity, learning new things, the pleasure of conducting some research and social engagement activities or incentive mechanisms, like payment or various rewards (Palacin et al. 2021).

Despite the large-scale usage of mobile apps, the transmission of information for various socio-economic purposes is conditioned by the economic level of that community. In Laikipia, a community in Kenya, the transmission of information on ticks and tick-borne diseases depends on the digitalization degree and the existing technological connection. Life quality is better where there are mobile phones that may be used for economic activities (Chepkwony et al. 2018). The same idea was supported also by Haddad and de Nazelle (2018), as their study participants considered that technology played a significant role in increasing their awareness on air pollution.

Although there are still a series of uncertainties regarding the effects of VGI, dedicated mobile apps may surprise the perceptions of various subjects and even different perceptions towards other sociological investigations (Gundelund et al. 2022; Schrammeijer et al. 2022). However, there are drawbacks in terms of the usability of such apps. For this reason, in case of a tested app such as We Comment!, the participants suggested the improvement of these mobile tools by paying attention to intuitive tool functions and the socio-demographic details of the respondents (Garcia et al. 2020). Also, the feedback provided to the participants involved in these participatory efforts is very important, although often considered a limitation, it has the potential for future development (Barrie et al. 2019; Beatty et al. 2021; Lowry et al. 2019; MacPhail et al. 2020). Equally important is the selection of an adequate communication strategy that effectively reaches the target audience (Martin et al. 2021; Nieman et al. 2020).

The evolution of the landscape of mobile apps in CS (Fig. 4) should be interpreted with caution, conditioned by the short time interval of 5 years. The number of the studies published in any of the 5 years of the considered interval ranges between 0 and 20. The highest values are recorded for Terrestrial species monitoring, regardless of the year. The analyzed data reveal two types of subdomains: a) subdomain where the use of mobile app in CS is continuous (studies identified in at least 3 consecutive years); b) subdomain where the use of mobile apps in CS is scarce.

Fig. 4: The evolution of CS apps over time.
figure 4

Number of papers on different CS app domains and subdomains each year during 2018–2022.

Full size image

The first category of subdomains reflects the impact of the Covid-19 sanitary crisis. This means either the decrease, sometimes to zero, of the studies published after 2020, or the increase of their number. Hence, some subdomains like Water, Earthquakes or Floods were negatively impacted, while others like Managing epidemics, Population health monitoring or Zoonotic diseases revealed a significant growth.

Discussing the future of CS apps

Having analyzed the current domains and categories of apps used in citizen science, we now turn to a discussion of trends we consider worth addressing in the future to maximize the impact of apps on CS. Based on the observation of the growth trends until now we have reasons to postulate that they would also experience a similar development in the future. By critically engaging with these observed trends, we argue for the need to pay increasing attention to the following implications of these developments. These include enabling CS project environments, improving data collection tools, ensuring data quality, and promoting citizen engagement (Fig. 5).

Fig. 5: Future trends of CS apps.
figure 5

Figure 5 highlights possible trends of mobile apps use in citizen science we consider worth addressing in the future.

Full size image

In what follows we discuss each of the implications outlined in Fig. 5 in turn.

Systemizing citizen science contributions to open science

A global policy for a digital CS may become, in its turn, a viable solution for various prevention systems, such as public health management (Batsaikhan et al. 2020; Katapally 2020). It is obvious that the future applicability of CS apps shall be influenced by the efficiency of the scientific community to organize and disseminate information pertaining to the progress made in this field. In this regard, adhering to the FAIR principles’ characteristic to Data Science and Open Science is imperative for CS projects (Batsaikhan et al. 2020; Roman et al. 2021). Adhering to the above-mentioned principles in projects will address the limitation highlighted by Fischer et al. (2021), which pertains to providing access to CS project data to scientists not directly involved in the project. Literature reviews as independent studies on CS apps should also adhere to the FAIR principles, making possible the continuation of reviews by authors different from those in the initial team. We consider that the results will enable a more complex image on the targeted phenomena over longer periods of time, as well as the validation of specific search expressions in the databases.

Finding data protection solutions to truly enable CS as a gamechanger in social and humanistic sciences

The CS projects in social and humanistic sciences are poorly represented in the analyzed literature, not only because they are published in journals outside WoS (Kullenberg and Kasperowski, 2016), although this is also a relevant aspect. The most recent and tentative application of CS in the above-mentioned domains is also influenced by concerns related to the sensitivity of personal data being processed. To respond to this need for increased personal data protection, we expect to witness more studies addressing various risks associated with CS projects.

Exploring new technologies to increase CS performance

The implementation of new technologies that are cost-effective and tailored to different types of mobile devices can lead to increased accessibility and affordability of lab analyses in CS projects. This, in turn, can expand the spatial and temporal coverage of monitoring campaigns and democratize information, making CS more inclusive and participatory (Bshouty et al. 2019; Chen et al. 2020; Costa et al. 2020). One of the CS tools that has achieved an extended reach and has become a significant catalyst is Digital Earth, described as “a participation model of our world” (Brovelli et al. 2020). We argue in favor of Digital Earth as a globally interconnected assembly of technologies, based on the principle that it is richer than the sum of its parts. Dedicated solutions for the data automation may contribute to the improvement of the data validation process (Bibrzycki et al. 2020; Ožana et al. 2019; Schade et al. 2019). Identification of other types of data, such as satellite or RPAS (Remotely Piloted Aircraft System) remote sensing, represent useful directions to investigate (Annis and Nardi 2019; Laso Bayas et al. 2020; Theuerkauf et al. 2022).

Designing and adopting large and reconfigurable platforms

Bonney et al.’s (2016) statement that many CS projects are not really “new” is represented in detail in Fig. 3, where we can observe that some CS projects collect similar data. This can confuse the group of potential participants and results in various irregular data sets instead of contributing to large and useful data sets. Hence, the proposed solution remains relevant to reduce project redundancy by fostering collaboration with existing proven projects and adapting them to the specific area and topic of interest. The technical variant of the solution could be a reusable and reconfigurable platform for deploying and monitoring collaborative participatory sensing projects (David Izquierdo-Gómez 2022; Moustard et al. 2021; Zaman et al. 2021). The open-source software variant, even if it reduces the costs of the project, does not eliminate for now the previously mentioned drawbacks (Anshori et al. 2022; Fraisl et al. 2022). For this to happen, connection to the same database is necessary, as well as its management by experts capable of processing it.

Assessing CS apps and their available functionalities

Indeed, it has been observed that CS apps may not always utilize all available functionalities that could potentially maximize their efficiency. As such, there is a need for more studies to assess the overall effectiveness of CS apps across various topics (Howard et al. 2022) or specific themes such as, for example, the design perspective (Huang and Benyoucef 2022). Furthermore, literature reviews based on the available functionalities of CS apps can serve as a driving force for more alert progress in the field.

Designing platforms for topic/theme groups

Mitigating redundancy, focusing volunteers’ efforts, and utilizing global databases with higher applicability will contribute to the grouping of apps based on complementary topics. It is already the case of some projects such as the GLOBE Program, with GLOBE Observer app (Low et al. 2021) or those gathering information about multiple types of disasters (Jacobs et al. 2019). In addition, a guideline that can be generalized and scaled for each of the categories of the group needs to be made available to citizens, regardless of their engagement level (Paul et al. 2021). The resources saved in such manner can be invested in the ongoing improvement of CS platforms and tools and their availability in as many languages as possible. Such fortunate cases may be imagined only if they are based on global networks composed of scientists and other stakeholders.

Understanding CS contribution to scientific literacy

Among the most common expectations from CS is a significant contribution to scientific literacy among the population, starting with the volunteers involved. We expect that many impact studies find ways to assess this contribution, on the one hand, while on the other hand, identify solutions to maximize the impact. Environmental literacy and health literacy are relevant examples in this regard (Altrudi 2021; Díaz et al. 2021). The integration of CS projects in secondary schools leads to an early engagement of individuals in these participatory efforts (Echeverria et al. 2021).

Capturing traditional knowledge

The prospects of collecting and managing traditional knowledge into large databases are now more feasible, offering potential for preserving a part of the cultural heritage of communities. Moreover, any intervention could be adjusted depending on the indigenous knowledge regardless of whether it relates or not to environmental protection (Kipp et al. 2019), agricultural practices (Laso Bayas et al. 2020; Thompson et al. 2019) etc.

Engaging citizen scientists to ensure project success

The good level of engagement in CS projects is influenced by the trust they have in the project initiators (Beza et al. 2018) and is often driven by non-pecuniary benefits rather than by financial compensation (Jacobs et al. 2019). Understanding technological, cultural, and environmental motivators explaining the adoption of CS apps (Cardoso-Andrade et al. 2022; Kowler et al. 2020) become common objectives. The engagement of volunteers in the research activity does not end with the completion of data collection. For this reason, many authors suggest that it is necessary to involve them in the validation of their own data. This could contribute to the increased loyalty of the volunteers involved in CS projects (Barrie et al. 2019). Also, the insertion of some minimal thresholds for the number of participants necessary for the validation of CS projects is another future research direction (Gundelund et al. 2021). Longitudinal studies reveal that social constructs formed at a specific point in time are subject to transformation and evolution over time (Beza et al. 2018). Furthermore, data that is collected and proves to be valuable not only to stakeholders, but also to the public, serves as an additional incentive for citizen engagement in their local communities (Ganzleben and Kazmierczak 2020).

Conclusions

The advancement of ITCs greatly benefitted CS so far, as we observed in our literature review. Through specific spatial-temporal data collection and visualization tools, citizens are empowered not just to become active contributors to science, but also to tackle issues they are passionate about, be it about reducing risk and pollution, conserving biodiversity or improving well-being. Crowd-sensed and crowd-sourced data collected by citizen scientists complements scientists’ and decision-makers’ understanding of specific issues, potentially improving scientific progress and policymaking. Thus, we argue, that the CS apps described in our literature review present promising attributes to complement the digital agora with spatial-temporal knowledge. First, it enables direct participation in scientific processes through virtual spaces. With smartphones and GIS technologies becoming increasingly part of the life of many citizens, CS apps make geoparticipation easier than ever before. There are still segments of the population that have no access to the necessary technological connection to participate. However, participation in CS through apps is nevertheless diverse. Some of the CS apps reviewed in this study were designed with specific target audiences in mind, indicating that people with different socio-demographic backgrounds defined by variables such as age, vocation or specific interests and hobbies are able and willing to engage. Future inquiries into the range and extent of participation in CS by different population segments could result in improvements in CS app functionality and citizen scientist engagement to ensure project success. CS apps also support different degrees of involvement in decision-making, from contributing with data collection to better understanding the issues set forth by scientists or decision-makers to framing community concerns through citizen-led CS projects. Lastly, while CS apps are not necessarily used to further a dialog between citizens, scientists and policymakers, it allows them to cooperate in spatial data collection and presentation. This is important in framing issues relevant for communities at any scale and in laying the ground for social change, be it improvements to city management, development of new policies, furthering scientific literacy or even encouraging behavioral change.

A further consideration worthy of note regards the possible future trends of apps in citizen science. We identified three content categories of the existing literature including the specific monitoring tools used for CS, the associated techniques and methods applied and the participatory engagement of citizen scientists. We expect these categories to be further explored in the future along with others, such as enabling a prolific and safe environment for citizen science projects and improving tools and methods as well as data quality. Not many of the studies we reviewed already follow the FAIR principles, characteristic to open science. However, in recent years the FAIR principles have become a hot topic in the scientific community. We find this tendency also promising for CS because it will allow the scientific community, citizens and citizen scientists alike, to improve CS apps by making use of already existent tools or learning from the experiences of others.

As for our study design, the most evident limitation is the exclusive identification of projects with scientific results published in peer reviewed journals with JCRs’ 5-Year impact. This led to a poor representation of Social Sciences, and Arts & Humanities, putting at a disadvantage the studies published in journals outside WoS. The results of this paper are nonetheless promising, revealing the increased and steadily expanding potential of VGI use in the development of citizen science. The contribution of the paper in terms of methodological approach of research synthesis lies in encouraging the use of the recursive searches method integrated within the PRISMA flow, with significantly better results in the field of CS literature reviews. The further testing of this method in CS literature analysis could target more specific segments of using mobile apps in CS or other emerging concepts and methods.