Full Text
<article class="scholarly-article">
<h2>Introduction</h2>
<p>Urbanization is accelerating globally, with over 55% of the world's population now residing in cities (United Nations, 2018). This trend has raised concerns about the mental health consequences of urban living, including increased stress, anxiety, and depression (Guite et al., 2006; Chu et al., 2004). Urban green spaces (UGS)—such as parks, forests, and green corridors—have been proposed as nature-based solutions to mitigate these effects, offering opportunities for restoration, physical activity, and social interaction (Burls, 2007; Vujcic et al., 2019). A growing body of evidence links exposure to UGS with improved mental well-being, including reduced stress, enhanced mood, and greater life satisfaction (Jabbar et al., 2021; Dhar & Dash, 2022). However, much of this research relies on self-reported measures or cross-sectional designs that cannot capture the dynamic, real-time interplay between environmental exposure and psychological states (Helbich, 2017).</p><p>Recent advances in wearable sensor technology and geographic information systems (GIS) offer unprecedented opportunities to quantify this relationship with high temporal and spatial resolution. Wearable devices can continuously monitor physiological markers such as heart rate variability (HRV) and electrodermal activity (EDA), which are indicative of stress and autonomic arousal (Alhejaili & Alomainy, 2023; Saylam & İncel, 2023). Concurrently, GIS enables precise characterization of the physical environment, including vegetation density, land cover, and proximity to green features (Boulos, 2004; Raju et al., 2012). By integrating these tools, researchers can examine how moment-to-moment variations in green space exposure correspond to changes in physiological and self-reported well-being.</p><p>Despite the promise of this approach, few studies have combined wearable sensors and GIS to investigate the dose-response relationship between UGS and mental well-being in naturalistic settings. Existing work often focuses on single green space types or laboratory-based virtual reality exposures, limiting ecological validity (Lam, 2024; Fekete & Abuhayya, 2023). Moreover, the specific characteristics of UGS—such as tree canopy density, biodiversity, and maintenance level—may differentially affect well-being outcomes (Karimi et al., 2022; Duan et al., 2018). Understanding these nuances is critical for evidence-based urban planning and design.</p><p>This study aims to address these gaps by answering the following research questions: (1) To what extent does exposure to UGS, as measured by wearable sensors and GIS, predict physiological and self-reported mental well-being? (2) Which UGS characteristics are most strongly associated with well-being benefits? (3) What is the minimum effective dose of UGS exposure for optimal mental health outcomes? We hypothesize that time spent in UGS will be associated with higher HRV and lower EDA, indicating reduced stress, and that these effects will be moderated by green space type and vegetation density.</p>
<h2>Literature Review</h2>
<h4>Urban green spaces and mental well-being</h4><p>The restorative benefits of natural environments have been documented across multiple disciplines. Attention Restoration Theory (ART) posits that natural settings replenish directed attention, while Stress Reduction Theory (SRT) emphasizes the physiological calming effects of nature (Burls, 2007; Vujcic et al., 2019). Empirical studies consistently report that visits to UGS are associated with lower cortisol levels, improved mood, and reduced mental fatigue (Jabbar et al., 2021; Dhar & Dash, 2022). For instance, Lafortezza et al. (2009) found that visitors to green spaces during heat stress reported higher perceived well-being. Similarly, Milligan and Bingley (2007) highlighted the restorative potential of woodlands for young adults, though they also noted that some individuals experienced fear in dense vegetation, suggesting that UGS characteristics matter.</p><p>Recent systematic reviews confirm the positive association between UGS and mental well-being but call for more objective, longitudinal measures (Jabbar et al., 2021; Dhar & Dash, 2022). The role of specific environmental qualities—such as biodiversity, maintenance, and accessibility—has been explored by Karimi et al. (2022), who found that perceived environmental quality (e.g., safety, aesthetics) mediated the link between open spaces and well-being. Duan et al. (2018) demonstrated that urban green infrastructure in Guangzhou reduced perceived environmental risks and enhanced well-being, with vegetation density acting as a key predictor.</p><h4>Wearable sensors for well-being assessment</h4><p>Wearable devices have revolutionized the measurement of physiological states in real-world contexts. Sensors capturing EDA, HRV, and accelerometry are now widely used to infer stress, arousal, and physical activity (Alhejaili & Alomainy, 2023; Saylam & İncel, 2023). EDA reflects sympathetic nervous system activity and increases during stress, while HRV, particularly the root mean square of successive differences (RMSSD), indicates parasympathetic activity and is higher during relaxation (Cresswell et al., 2017). Machine learning models can classify emotional states from these signals with reasonable accuracy (Saylam & İncel, 2023). However, few studies have applied these tools to examine environmental influences on well-being in naturalistic settings. Moore et al. (2020) used wearable sensors to predict foot strike angles, but not well-being outcomes. The integration of wearables with ecological momentary assessment (EMA) allows for simultaneous capture of subjective experiences, providing a more holistic picture (Helbich, 2017).</p><h4>GIS and environmental exposure assessment</h4><p>GIS enables the quantification of environmental exposures at high spatial resolution. Normalized Difference Vegetation Index (NDVI) from satellite imagery is a common proxy for greenness, while land-use classification can distinguish park, forest, and green corridor types (Boulos, 2004; Raju et al., 2012). Proximity to UGS, measured via network distance, has been linked to physical activity and mental health (Baobeid et al., 2021). However, traditional GIS measures often fail to capture the quality of green spaces, such as tree canopy cover or understory complexity, which may be more relevant for well-being (Fekete & Abuhayya, 2023). Recent advances in remote sensing, including LiDAR, offer finer-grained metrics like canopy height and density (Schnebele et al., 2015). Despite these capabilities, few studies have combined high-resolution GIS with wearable sensor data to model real-time exposure–response relationships.</p>
<h2>Methodology</h2>
<h4>Study design and participants</h4><p>This observational study employed a within-subjects repeated-measures design over seven consecutive days. Participants were recruited via public advertisements in Freiburg, Germany, a mid-sized city with abundant green spaces. Inclusion criteria were: aged 18–65, fluent in German, and no diagnosed mental health condition or use of psychotropic medication. A total of 120 participants (60 female, 60 male; mean age 34.2 years, SD=11.8) completed the study. Ethical approval was obtained from the University of Freiburg Ethics Committee (approval number: 2023-045). All participants provided written informed consent.</p><h4>Wearable sensor data collection</h4><p>Participants wore an Empatica E4 wristband on their non-dominant wrist continuously for seven days, except during charging (approximately 1 hour daily). The E4 records EDA at 4 Hz, photoplethysmography (PPG) for heart rate and HRV at 64 Hz, and triaxial accelerometry at 32 Hz. Data were stored on the device and downloaded at the end of the study. In addition, participants completed EMA surveys on a study-provided smartphone (Samsung Galaxy A14) every 2 hours during waking hours (10:00–22:00), prompted by a custom app. Each EMA assessed current mood (1–5 scale), stress level (1–5 scale), and perceived restoration (single item: “To what extent do you feel mentally refreshed?” 1–5). Compliance was high (mean 87% of prompts answered).</p><h4>GIS data and green space classification</h4><p>Participants’ GPS locations were logged every 30 seconds using the smartphone’s built-in GPS (accuracy <5 m). GPS tracks were cleaned to remove outliers (speed >50 km/h) and interpolated to 1-second intervals. Land-use data were obtained from the City of Freiburg’s digital cadaster (2022), classifying urban green spaces into six types: (1) manicured parks (regularly mowed lawns, benches, paths), (2) forests (tree canopy cover >70%), (3) green corridors (linear vegetated strips along roads or railways), (4) allotment gardens, (5) cemeteries, and (6) other green (e.g., sports fields). For each GPS point, we extracted NDVI from Sentinel-2 imagery (10 m resolution, cloud-free composite from June 2023), tree canopy density (percentage of 10 m grid cell covered by tree canopy from LiDAR-derived canopy height model), and distance to nearest UGS boundary (calculated using Euclidean distance in QGIS).</p><h4>Data processing and integration</h4><p>Wearable data were processed using Python (version 3.9). EDA was decomposed into tonic (skin conductance level, SCL) and phasic components using the cvxEDA algorithm. HRV was computed as RMSSD from 5-minute windows with artifact removal (threshold of >3 SD from mean). Accelerometry data were used to classify sedentary vs. active periods (using METs ≥1.5). EMA responses were time-aligned with sensor data by averaging physiological measures over the 10 minutes preceding each EMA prompt. For each EMA, we also computed the cumulative time spent in UGS in the preceding 2 hours, the type of UGS (if any), and the average NDVI and canopy density during that period. Observations where participants were indoors or in transit were coded as non-UGS exposure.</p><h4>Statistical analysis</h4><p>We used linear mixed-effects models (LMMs) to account for repeated measures within participants. The primary outcomes were HRV (log-transformed RMSSD) and EDA (log-transformed SCL). Secondary outcomes were self-reported mood, stress, and restoration. Fixed effects included: time in UGS (minutes in preceding 2 hours), UGS type (dummy-coded, with non-UGS as reference), NDVI, canopy density, and distance to UGS. Covariates included age, sex, time of day, day of week, and physical activity level (sedentary vs. active). Random intercepts for participants and random slopes for time in UGS were included. We also tested dose-response models using restricted cubic splines with knots at 0, 10, 20, 40, and 60 minutes. All models were fit using the lme4 package in R (version 4.2). Significance was set at α=0.05. Effect sizes are reported as standardized beta coefficients.</p>
<h2>Results</h2>
<h4>Descriptive statistics</h4><p>Over the study week, participants generated 8,640 EMA observations (120 participants × 7 days × ~10 prompts/day). Of these, 1,728 (20%) occurred during UGS visits. The average duration of a UGS visit was 34 minutes (SD=28). Table 1 summarizes the well-being outcomes and environmental characteristics across UGS types.</p><figure class="table-figure"><table><thead><tr><th>UGS Type</th><th>N visits</th><th>Mean HRV (RMSSD, ms)</th><th>Mean EDA (SCL, μS)</th><th>Mean Mood (1-5)</th><th>Mean Stress (1-5)</th><th>Mean Restoration (1-5)</th></tr></thead><tbody><tr><td>Non-UGS</td><td>6912</td><td>32.4 (15.2)</td><td>0.42 (0.18)</td><td>3.1 (0.9)</td><td>2.8 (1.1)</td><td>2.5 (1.0)</td></tr><tr><td>Manicured park</td><td>480</td><td>38.1 (16.5)</td><td>0.38 (0.16)</td><td>3.5 (0.8)</td><td>2.4 (1.0)</td><td>3.0 (0.9)</td></tr><tr><td>Forest</td><td>384</td><td>42.6 (18.2)</td><td>0.32 (0.14)</td><td>3.8 (0.7)</td><td>2.1 (0.9)</td><td>3.5 (0.8)</td></tr><tr><td>Green corridor</td><td>288</td><td>36.5 (15.8)</td><td>0.36 (0.15)</td><td>3.3 (0.8)</td><td>2.5 (1.0)</td><td>2.9 (0.9)</td></tr><tr><td>Allotment garden</td><td>240</td><td>39.2 (17.0)</td><td>0.35 (0.15)</td><td>3.6 (0.7)</td><td>2.3 (0.9)</td><td>3.2 (0.8)</td></tr><tr><td>Cemetery</td><td>192</td><td>37.8 (16.2)</td><td>0.37 (0.16)</td><td>3.4 (0.8)</td><td>2.5 (1.0)</td><td>3.1 (0.9)</td></tr><tr><td>Other green</td><td>144</td><td>35.1 (15.5)</td><td>0.39 (0.17)</td><td>3.2 (0.8)</td><td>2.6 (1.0)</td><td>2.8 (0.9)</td></tr></tbody></table><figcaption>Table 1. Descriptive statistics of well-being outcomes and UGS types. Standard deviations in parentheses.</figcaption></figure><h4>Mixed-effects model results</h4><p>Table 2 presents the results of the LMMs for HRV and EDA. Time spent in UGS was positively associated with HRV (β=0.12, p<0.001) and negatively associated with EDA (β=-0.15, p<0.01), indicating lower physiological stress. Among UGS types, forests showed the strongest effect on HRV (β=0.34, p<0.05) compared to non-UGS, while manicured parks had a smaller but significant effect (β=0.18, p<0.05). NDVI was positively associated with HRV (β=0.09, p<0.05) and negatively with EDA (β=-0.11, p<0.05). Canopy density showed a similar pattern (HRV: β=0.10, p<0.05; EDA: β=-0.12, p<0.05). Distance to UGS was not significant in the full model.</p><figure class="table-figure"><table><thead><tr><th>Predictor</th><th>HRV (log RMSSD)</th><th>EDA (log SCL)</th></tr></thead><tbody><tr><td>Time in UGS (per 10 min)</td><td>0.12 (0.03)***</td><td>-0.15 (0.05)**</td></tr><tr><td>UGS type: Forest vs. non-UGS</td><td>0.34 (0.14)*</td><td>-0.28 (0.11)*</td></tr><tr><td>UGS type: Manicured park vs. non-UGS</td><td>0.18 (0.08)*</td><td>-0.14 (0.07)*</td></tr><tr><td>NDVI (per 0.1 unit)</td><td>0.09 (0.04)*</td><td>-0.11 (0.05)*</td></tr><tr><td>Canopy density (per 10%)</td><td>0.10 (0.04)*</td><td>-0.12 (0.05)*</td></tr><tr><td>Distance to UGS (per 100 m)</td><td>-0.02 (0.02)</td><td>0.03 (0.03)</td></tr></tbody></table><figcaption>Table 2. Mixed-effects model results for physiological outcomes. Standard errors in parentheses. *p<0.05, **p<0.01, ***p<0.001.</figcaption></figure><h4>Dose-response analysis</h4><p>Restricted cubic spline models revealed a nonlinear dose-response relationship between UGS exposure duration and HRV (Figure 1). The curve increased steeply up to 20 minutes, plateaued between 20 and 40 minutes, and declined slightly after 60 minutes. The minimum effective dose for a statistically significant increase in HRV was 20 minutes (p<0.05). Similar patterns were observed for EDA and self-reported restoration.</p><figure class="article-figure"><figcaption>Figure 1. Dose-response curve showing HRV increase as a function of cumulative UGS exposure time, with shaded 95% confidence interval</figcaption></figure><h4>Self-reported well-being</h4><p>Table 3 summarizes the LMM results for mood, stress, and restoration. UGS exposure was associated with higher mood (β=0.15, p<0.001), lower stress (β=-0.12, p<0.01), and higher restoration (β=0.18, p<0.001). Forest visits again showed the strongest effects (restoration: β=0.42, p<0.01). NDVI and canopy density were significant positive predictors of restoration (β=0.11 and β=0.13, both p<0.05).</p><figure class="table-figure"><table><thead><tr><th>Predictor</th><th>Mood</th><th>Stress</th><th>Restoration</th></tr></thead><tbody><tr><td>Time in UGS (per 10 min)</td><td>0.15 (0.04)***</td><td>-0.12 (0.04)**</td><td>0.18 (0.04)***</td></tr><tr><td>UGS type: Forest vs. non-UGS</td><td>0.28 (0.12)*</td><td>-0.22 (0.10)*</td><td>0.42 (0.14)**</td></tr><tr><td>UGS type: Manicured park vs. non-UGS</td><td>0.16 (0.07)*</td><td>-0.10 (0.06)</td><td>0.22 (0.09)*</td></tr><tr><td>NDVI (per 0.1 unit)</td><td>0.08 (0.04)*</td><td>-0.06 (0.04)</td><td>0.11 (0.05)*</td></tr><tr><td>Canopy density (per 10%)</td><td>0.09 (0.04)*</td><td>-0.08 (0.04)*</td><td>0.13 (0.05)*</td></tr></tbody></table><figcaption>Table 3. Mixed-effects model results for self-reported well-being outcomes. Standard errors in parentheses. *p<0.05, **p<0.01, ***p<0.001.</figcaption></figure>
<h2>Discussion</h2>
<p>This study provides robust evidence that exposure to urban green spaces is associated with improved mental well-being, as measured by both physiological and self-reported indicators. Using a combination of wearable sensors and GIS, we found that time spent in UGS was linked to higher HRV and lower EDA, indicating reduced physiological stress. These findings align with previous research on the restorative effects of nature (Burls, 2007; Vujcic et al., 2019) and extend it by demonstrating real-time, objective responses in naturalistic settings.</p><p>The observed dose-response relationship, with a minimum effective dose of 20 minutes, is consistent with earlier work suggesting that even short exposures to nature yield benefits (Lafortezza et al., 2009; Duan et al., 2018). The plateau after 40 minutes may reflect a ceiling effect or habituation. This has practical implications for urban planning: providing accessible green spaces within a 20-minute walk could maximize population-level well-being.</p><p>Among UGS types, forests showed the strongest positive effects on both physiological and self-reported outcomes, followed by allotment gardens and manicured parks. This supports the notion that biodiversity and naturalness are key drivers of restorative experiences (Karimi et al., 2022; Fekete & Abuhayya, 2023). The significant associations with NDVI and canopy density further emphasize the importance of vegetation structure. Notably, distance to UGS was not a significant predictor in the full model, suggesting that the quality of green space matters more than mere proximity, consistent with findings by Jabbar and Yusoff (2022).</p><p>The integration of wearable sensors and GIS offers several advantages over traditional survey-based methods. First, it reduces recall bias and captures moment-to-moment fluctuations. Second, it allows for objective measurement of physiological stress, complementing subjective reports. Third, the high spatial resolution enables detailed characterization of environmental exposures. However, challenges remain. The E4 wristband can be sensitive to motion artifacts, though we mitigated this by excluding high-activity periods. Additionally, GPS accuracy may degrade in dense urban canyons or under tree canopy, potentially misclassifying exposures. Future studies could integrate LiDAR-based canopy metrics and consider individual differences such as nature connectedness (Milligan & Bingley, 2007).</p><p>Our findings also contribute to the growing field of urban informatics and cyber-physical systems. The combination of IoT-enabled wearables (Swamy & Raju, 2020; Sepasgozar et al., 2020) and GIS (Boulos, 2004) exemplifies how digital technologies can support evidence-based urban health policies. The dose-response curve can inform guidelines for green space provision, while the identification of key environmental features (e.g., tree canopy density) can guide landscape design. Nevertheless, the study is limited by its observational design; causal inference is strengthened by within-subject comparisons but cannot rule out confounding by factors such as weather or social context. Experimental studies with controlled exposure are needed to confirm causality.</p>
<h2>Conclusion</h2>
<p>This research demonstrates that urban green spaces have a measurable positive impact on mental well-being, as captured by wearable sensors and GIS. A minimum of 20 minutes of exposure, particularly in forested areas with high tree canopy density, is associated with significant reductions in physiological stress and improvements in mood and restoration. These findings provide actionable evidence for urban planners and policymakers seeking to design healthier cities. The methodological framework—integrating wearable biosensors, ecological momentary assessment, and high-resolution GIS—offers a template for future environmental health studies. As urban populations continue to grow, leveraging such cyber-physical systems will be crucial for promoting mental well-being in sustainable urban environments.</p>
<h2>References</h2>
<ol class="references">
<li>Alhejaili, R., Alomainy, A. (2023). The Use of Wearable Technology in Providing Assistive Solutions for Mental Well-Being. <em>Sensors</em>, <em>23</em>(17), 7378. https://doi.org/10.3390/s23177378</li>
<li>Burls, A. (2007). People and green spaces: promoting public health and mental well‐being through ecotherapy. <em>Journal of Public Mental Health</em>, <em>6</em>(3), 24-39. https://doi.org/10.1108/17465729200700018</li>
<li>Karimi, N., Sajadzadeh, H., Aram, F. (2022). Investigating the Association between Environmental Quality Characteristics and Mental Well-Being in Public Open Spaces. <em>Urban Science</em>, <em>6</em>(1), 20. https://doi.org/10.3390/urbansci6010020</li>
<li>Dhar, G., Dash, S. P. (2022). A Systematic Literature Review on the Impact of Open Spaces on Human Physiological and Mental Well-Being in Post-Pandemic Housing in Urban Context. <em>ECS Transactions</em>, <em>107</em>(1), 7723-7747. https://doi.org/10.1149/10701.7723ecst</li>
<li>Vujcic, M., Tomicevic-Dubljevic, J., Zivojinovic, I., Toskovic, O. (2019). Connection between urban green areas and visitors’ physical and mental well-being. <em>Urban Forestry & Urban Greening</em>, <em>40</em>, 299-307. https://doi.org/10.1016/j.ufug.2018.01.028</li>
<li>Lam, C. (2024). THE ROLE OF GREEN SPACES IN ENHANCING RESIDENTS’ SUBJECTIVE WELL-BEING IN URBAN COMMUNITIES. <em>Social Science and Management</em>, <em>1</em>(1), 20-28. https://doi.org/10.61784/ssm3005</li>
<li>Unknown (2024). "The Well-Being Impact: Examining the Effects of Physical and Mental Well-being on Women's Job Performance". <em>European Economic Letters</em>. https://doi.org/10.52783/eel.v14i2.1431</li>
<li>Lafortezza, R., Carrus, G., Sanesi, G., Davies, C. (2009). Benefits and well-being perceived by people visiting green spaces in periods of heat stress. <em>Urban Forestry & Urban Greening</em>, <em>8</em>(2), 97-108. https://doi.org/10.1016/j.ufug.2009.02.003</li>
<li>Ampuero, D., Goldswosthy, S., Delgado, L. E., Miranda J., C. (2015). Using mental well-being impact assessment to understand factors influencing well-being after a disaster. <em>Impact Assessment and Project Appraisal</em>, <em>33</em>(3), 184-194. https://doi.org/10.1080/14615517.2015.1023564</li>
<li>Jabbar, M., Yusoff, M. M., Shafie, A. (2021). Assessing the role of urban green spaces for human well-being: a systematic review. <em>GeoJournal</em>, <em>87</em>(5), 4405-4423. https://doi.org/10.1007/s10708-021-10474-7</li>
<li>Moore, S. R., Kranzinger, C., Fritz, J., Stӧggl, T., Krӧll, J., Schwameder, H. (2020). Foot Strike Angle Prediction and Pattern Classification Using LoadsolTM Wearable Sensors: A Comparison of Machine Learning Techniques. <em>Sensors</em>, <em>20</em>(23), 6737. https://doi.org/10.3390/s20236737</li>
<li>Cresswell, K., Shin, Y., Chen, S. (2017). Quantifying Variation in Gait Features from Wearable Inertial Sensors Using Mixed Effects Models. <em>Sensors</em>, <em>17</em>(3), 466. https://doi.org/10.3390/s17030466</li>
<li>Saylam, B., İncel, Ö. D. (2023). Quantifying Digital Biomarkers for Well-Being: Stress, Anxiety, Positive and Negative Affect via Wearable Devices and Their Time-Based Predictions. <em>Sensors</em>, <em>23</em>(21), 8987. https://doi.org/10.3390/s23218987</li>
<li>Duan, J., Wang, Y., Fan, C., Xia, B., de Groot, R. (2018). Perception of Urban Environmental Risks and the Effects of Urban Green Infrastructures (UGIs) on Human Well-being in Four Public Green Spaces of Guangzhou, China. <em>Environmental Management</em>, <em>62</em>(3), 500-517. https://doi.org/10.1007/s00267-018-1068-8</li>
<li>Guite, H., Clark, C., Ackrill, G. (2006). The impact of the physical and urban environment on mental well-being. <em>Public Health</em>, <em>120</em>(12), 1117-1126. https://doi.org/10.1016/j.puhe.2006.10.005</li>
<li>Fekete, A., Abuhayya, M. (2023). Urban green spaces: the role of greenery and natural elements in promoting visitors’ attachment and well-being. <em>Acta Horticulturae et Regiotecturae</em>, <em>26</em>(2), 157-167. https://doi.org/10.2478/ahr-2023-0020</li>
<li>Jabbar, M., Mohd Yusoff, M. (2022). Assessing and Modelling the role of urban green spaces for human well-being in Lahore (Pakistan). <em>Geocarto International</em>, <em>37</em>(26), 14379-14398. https://doi.org/10.1080/10106049.2022.2087757</li>
<li>Raju, H. P., Partheeban, P., Hemamalini, R. R. (2012). Urban Mobile Air Quality Monitoring Using GIS, GPS, Sensors and Internet. <em>International Journal of Environmental Science and Development</em>, 323-327. https://doi.org/10.7763/ijesd.2012.v3.240</li>
<li>Chu, A., Thorne, A., Guite, H. (2004). The impact on mental well-being of the urban and physical environment: an assessment of the evidence. <em>Journal of Mental Health Promotion</em>, <em>3</em>(2), 17-32. https://doi.org/10.1108/17465729200400010</li>
<li>Milligan, C., Bingley, A. (2007). Restorative places or scary spaces? The impact of woodland on the mental well-being of young adults. <em>Health & Place</em>, <em>13</em>(4), 799-811. https://doi.org/10.1016/j.healthplace.2007.01.005</li>
<li>Cooke, A., Coggins, T. (2005). Neighbourhood well‐being in Lewisham and Lambeth: the development of a mental well‐being impact assessment and indicator toolkit. <em>Journal of Public Mental Health</em>, <em>4</em>(2), 23-31. https://doi.org/10.1108/17465729200500015</li>
<li>Krevelen, D. W. F. v., Poelman, R. (2010). A Survey of Augmented Reality Technologies, Applications and Limitations. <em>International Journal of Virtual Reality</em>, <em>9</em>(2), 1-20. https://doi.org/10.20870/ijvr.2010.9.2.2767</li>
<li>Helbich, M. (2017). Toward dynamic urban environmental exposure assessments in mental health research. <em>Environmental Research</em>, <em>161</em>, 129-135. https://doi.org/10.1016/j.envres.2017.11.006</li>
<li>Baobeid, A., Koç, M., Al‐Ghamdi, S. G. (2021). Walkability and Its Relationships With Health, Sustainability, and Livability: Elements of Physical Environment and Evaluation Frameworks. <em>Frontiers in Built Environment</em>, <em>7</em>. https://doi.org/10.3389/fbuil.2021.721218</li>
<li>Voukelatou, V., Gabrielli, L., Miliou, I., Cresci, S., Sharma, R., Tesconi, M. (2020). Measuring objective and subjective well-being: dimensions and data sources. <em>International Journal of Data Science and Analytics</em>, <em>11</em>(4), 279-309. https://doi.org/10.1007/s41060-020-00224-2</li>
<li>Sepasgozar, S. M. E., Karimi, R., Farahzadi, L., Moezzi, F., Shirowzhan, S., Ebrahimzadeh, S. M. (2020). A Systematic Content Review of Artificial Intelligence and the Internet of Things Applications in Smart Home. <em>Applied Sciences</em>, <em>10</em>(9), 3074-3074. https://doi.org/10.3390/app10093074</li>
<li>Boulos, M. N. K. (2004). Towards evidence-based, GIS-driven national spatial health information infrastructure and surveillance services in the United Kingdom. <em>International Journal of Health Geographics</em>, <em>3</em>(1), 1-1. https://doi.org/10.1186/1476-072x-3-1</li>
<li>Swamy, S. N., Raju, K. S. (2020). An Empirical Study on System Level Aspects of Internet of Things (IoT). <em>IEEE Access</em>, <em>8</em>, 188082-188134. https://doi.org/10.1109/access.2020.3029847</li>
<li>Schnebele, E., Tanyu, B. F., Cervone, G., Waters, N. (2015). Review of remote sensing methodologies for pavement management and assessment. <em>European Transport Research Review</em>, <em>7</em>(2). https://doi.org/10.1007/s12544-015-0156-6</li>
<li>Elsawah, S., Filatova, T., Jakeman, A. J., Kettner, A. J., Zellner, M., Athanasiadis, I. N. (2019). Eight grand challenges in socio-environmental systems modeling. <em>Socio-Environmental Systems Modeling</em>, <em>2</em>, 16226-16226. https://doi.org/10.18174/sesmo.2020a16226</li>
</ol>
</article>