The Paravur aquifer is a natural laboratory for testing proximity-based vulnerability assumptions. The municipality covers 16.19 square kilometres, is surrounded on three sides by water — Paravur Lake, the Edava-Nadayara Kayal and the Ithikkara River — and hosts roughly 8,000 open dug wells, most reaching just 6 to 15 metres into a shallow unconfined alluvial aquifer. That aquifer consists of 3 to 8 metres of fine-to-medium sand containing 8 to 15 percent clay, predominantly montmorillonite-kaolinite assemblages, overlying laterite bedrock. Hydraulic conductivity of 5 to 15 metres per day allows seawater to migrate rapidly during the dry season, while the clay fraction sets the stage for cation exchange reactions that can chemically mask or mimic intrusion.

The research team divided the transect into three zones — Core (0–3 km), Mid (3–6 km) and Outer (6–8.7 km) — and collected 50, 25 and 25 samples respectively, allocating density proportionally to intrusion risk and well distribution. Every tenth sample was analysed in duplicate, yielding a mean relative percentage difference of 3.2 percent for major ions, and ionic balance errors fell within ±5 percent for all samples after re-analysis. Field parameters were measured with a multiparameter probe, and major ions were determined at an environmental monitoring laboratory following APHA standards, with samples stored at 4 °C and analysed within 48 hours.

Bulk salinity behaved conventionally. Mean total dissolved solids fell from 504.61 milligrams per litre in the Core Zone to 302.35 in the Mid Zone and 184.06 in the Outer Zone, a 63.5 percent reduction coast to inland. Electrical conductivity followed the same trajectory, from 354.18 to 123.32 microsiemens per centimetre. But pH ran in the opposite direction, rising from acidic coastal values of 5.76 to alkaline inland values of 7.27 — a 1.51-unit increase over 8.7 kilometres that contradicts the saline, alkaline signature seawater would normally impose at the shoreline. Statistical tests confirmed these gradients were genuine: Kruskal-Wallis and ANOVA analyses produced p-values below 0.001 for pH, sodium, conductivity, TDS and sulphate, with Cohen’s d effect sizes for Core-Outer comparisons reaching 2.54 for pH and 2.87 for sodium.

The process-based indicators told the inverted story. The chloride-to-sodium ratio, which in seawater stands at 1.81, increased systematically from 0.78 at the coast to 1.62 inland, with several inland wells within 10 percent of seawater stoichiometry. The chlorinity index climbed from 55.07 to 79.83 percent, crossing the threshold that indicates substantial seawater influence. Most strikingly, the Base Exchange Index, which distinguishes salinisation (negative values) from freshening (positive values), reversed from +2.01 meq/L in the Core Zone to −0.15 meq/L in the Outer Zone. Piper trilinear diagrams corroborated the pattern: coastal wells plotted as mixed, sulphate-enriched facies with acidic pH, while 88 percent of inland wells plotted squarely in the Na-Cl seawater field with alkaline pH and negligible sulphate.

Principal component analysis quantified the dual processes at work. On eight standardised variables, PC1 explained 52.3 percent of the variance with strong loadings on chloride (0.89), sodium (0.87) and electrical conductivity (0.83), a pattern consistent with conservative seawater mixing. PC1 scores declined coast to inland, mirroring the salinity gradient. PC2 explained 18.8 percent with a strong loading on BEX (0.81) alongside calcium (0.72) and magnesium (0.68), consistent with cation exchange. Mean PC2 scores shifted from positive at the coast to negative inland, and the calculated Exchange Contribution Index rose from 31.2 percent in the Core Zone to 42.7 percent in the Outer Zone — evidence that exchange reactions become the dominant control on water chemistry with distance from the sea.

The sulphate story adds an anthropogenic twist. Coastal wells averaged 123.53 milligrams per litre of sulphate, far above the 17.8 milligrams per litre expected from conservative seawater mixing at those chloride levels. With Paravur’s population density of roughly 6,214 persons per square kilometre and no centralised sewerage, domestic wastewater discharged through septic tanks and soak pits readily infiltrates the shallow aquifer, and sulphide oxidation and protein mineralisation in septic effluent are well-documented sulphate sources. The authors recommend isotopic analysis (δ34S and δ18O of sulphate) in future work to partition sources definitively, noting that nitrate measurements — absent from this study — would also have traced sewage contamination.

The irrigation assessment revealed what the authors call a freshening-contamination paradox. On conventional indices, the water looks broadly suitable: 96 percent of samples met the magnesium hazard criterion, 93 percent met the permeability index criterion, and 100 percent met the residual sodium carbonate criterion. Yet Kelly’s Ratio, which compares sodium to the sum of calcium and magnesium, exceeded the critical threshold of 1 in 91 percent of all samples — 100 percent in the Core Zone, 88 percent in the Mid Zone and 76 percent in the Outer Zone. Mean Kelly’s Ratio values of 2.15, 1.62 and 1.32 across the zones indicate sodium concentrations 1.3 to 2.2 times higher than divalent cations, a sodicity imbalance that threatens soil structure and permeability. Correlation analysis mechanistically linked forward cation exchange (negative PC2) to elevated sodium hazard (r = −0.58, p < 0.001).

What mechanism could deliver seawater signatures inland while the coast freshens? The authors propose three candidates, all requiring geophysical confirmation. Buried paleo-channels, mapped regionally by the Geological Survey of India, could act as high-permeability conduits bypassing the shoreline, analogous to documented systems in coastal Bangladesh and the Nile Delta. Variable laterite thickness of 2 to 6 metres could create differential vertical connectivity, with fractured inland laterite permitting upward seawater leakage. And semi-confined deeper horizons could enable density-driven inland migration, with well screens selectively sampling freshwater at the coast and saline water inland. Similar inverted patterns have been reported in the Mar del Plata aquifer in Argentina and the Gaza coastal aquifer, suggesting that a non-trivial share of the world’s heterogeneous coastal aquifers may behave this way.

The practical implications are stark. Monitoring networks concentrated near the shoreline, relying on electrical conductivity, TDS and chloride alone, would systematically miss the zones of greatest agricultural risk in aquifers like Paravur’s. The authors recommend extending networks inland regardless of measured salinity, coupling process-based indicators such as BEX and the chloride-to-sodium ratio with sodium-specific irrigation indices like Kelly’s Ratio, and refusing to classify low-salinity water with positive BEX as

Subject of Research: Assessing the dynamics and severity of seawater intrusion in a vulnerable coastal aquifer: an integrated hydrogeochemical facies and ionic ratios approach in Paravur, India

Article Title: Assessing the dynamics and severity of seawater intrusion in a vulnerable coastal aquifer: an integrated hydrogeochemical facies and ionic ratios approach in Paravur, India

Article References: Kani, M. K., Biju, G. N., & VS, V. (2026). Assessing the dynamics and severity of seawater intrusion in a vulnerable coastal aquifer: an integrated hydrogeochemical facies and ionic ratios approach in Paravur, India. Estuarine Management and Technologies, 3, 201-232. https://doi.org/10.3897/emt.3.198905

Image Credits: AI Generated

DOI: 10.3897/emt.3.198905

Keywords: Assessing, dynamics, severity, seawater, intrusion, vulnerable, coastal, aquifer, integrated, hydrogeochemical, facies, ionic

The study, published in Frontiers in Child and Adolescent Psychiatry, used a mixed-methods design that combined a quantitative survey with in-depth interviews. All participants completed the 20-item Center for Epidemiologic Studies Depression Scale, or CES-D, a widely validated screening tool scored from 0 to 60. Scores of 16 or above indicate clinically significant symptoms, and 49.68 percent of the students — 392 of 789 — crossed that threshold. The researchers then sorted scores into three ordered categories: low symptoms (0–15), mild-to-moderate symptoms (16–26), and severe symptoms (27–60), and used ordinal logistic regression to identify which of thirteen variables independently predicted movement up that severity ladder.

The statistical results were unambiguous. Compared with students reporting more than three hours of daily social-media use, those using it less than one hour per day had dramatically lower odds of greater symptom severity, with an adjusted odds ratio of 0.125 (95 percent confidence interval: 0.040–0.391). Students in the one-to-three-hour band also fared better, with an odds ratio of 0.424 (95 percent CI: 0.278–0.647). In raw numbers, only 4.55 percent of light users fell into the severe category, versus 24.11 percent of heavy users. Daily social-media use showed the strongest adjusted association of any factor in the model — stronger than academic year, major of study, grade point average, or source of personal support.

Yet when students rated eight potential stress domains on a four-point agreement scale, internet and social-media use ranked only fifth, endorsed by just 37.01 percent of respondents. Relationship and social environment topped the list at 56.40 percent agreement, followed by financial and economic status (53.87 percent) and the educational system (50.19 percent). This perception–association mismatch — the gap between what students perceive as stressful and what the statistics reveal — is the study’s central contribution, and it may explain why digital wellbeing is so often overlooked in university mental-health programs.

The qualitative strand of the study helps explain the blind spot. Twelve purposively selected students, six female and six male, aged 18 to 26, participated in semi-structured interviews covering social-media routines, online comparison, nighttime use and sleep, coping, and help-seeking. Thematic analysis revealed that social media was so deeply embedded in everyday student life — used after waking, during commutes, between classes, and before sleep for entertainment, news, communication, information seeking, academic tasks, and relief from boredom — that it was rarely identified as a source of stress. One participant captured the ambivalence: while using social media felt enjoyable, spending too long on it sometimes left a sense of having wasted time.

The interviews also revealed how the psychological impact of social media depends on patterns and interpretations of use rather than duration alone. Several students described upward social comparison when viewing posts about career success, travel, or desirable lifestyles, reporting feelings of pressure, inadequacy, or falling behind. One participant explained that seeing posts about good jobs or beautiful destinations sometimes triggered thoughts of not being good enough. But comparison was not universal. Another student consciously discounted online portrayals, noting that people usually post only the good things that happen to them, and therefore felt little pressure from others’ lives. Nighttime use emerged as a more consistent concern among heavier users, with late-night scrolling delaying sleep and producing fatigue, irritability, mental fog, and reduced concentration the next day — though some participants reported no sleep disturbance at all.

The third qualitative theme concerned how students manage emotional distress. Participants described a stepped pattern of coping: manageable distress was handled through self-directed strategies such as listening to music, drawing, watching videos, traveling, or eating enjoyable food, while social media itself could function as both a potential source of strain and a coping resource. When distress became harder to manage, close friends and family were approached first, and professional counseling was generally viewed as a later option. Reluctance to seek formal help was shaped by fear of judgment, privacy concerns, and kreng jai — the Thai cultural norm of not wanting to impose on others. This pattern aligns with the quantitative finding that students who confided in family (adjusted OR = 0.289, 95 percent CI: 0.161–0.519) or close friends (adjusted OR = 0.536, 95 percent CI: 0.309–0.929) had lower odds of greater symptom severity than those who consulted teachers or others.

Beyond social media, the regression model identified several other independent correlates of depressive-symptom severity. Students in the lowest grade point average band (2.00–2.49) had more than threefold higher odds of greater severity than those in the highest band (adjusted OR = 3.161, 95 percent CI: 1.933–5.165). Senior students showed higher odds than junior cohorts, and Health Sciences students had lower odds than the reference major (adjusted OR = 0.554, 95 percent CI: 0.384–0.801). Engineering and Technology students had the highest severe-symptom rate in bivariate comparisons at 35.66 percent, versus 11.59 percent for Health Sciences. No significant associations were found for sex, age, religion, hometown, smoking, alcohol use, or relationship status.

The authors are careful to note that these cross-sectional findings do not establish causality. Reverse or bidirectional relationships remain plausible: students experiencing more severe depressive symptoms may spend more time online for distraction, social connection, or coping, while prolonged or emotionally challenging online engagement may also worsen wellbeing. Unmeasured factors such as loneliness, sleep disturbance, academic stress, or pre-existing psychological distress could contribute to both greater social-media use and greater symptom severity. The study also measured social-media use only by daily duration, and the qualitative findings themselves demonstrate that similar duration can encompass very different psychological experiences. Convenience sampling at a single private Thai university further limits generalizability, and the CES-D is a screening instrument rather than a diagnostic interview.

Nevertheless, the perception–association mismatch carries practical weight for university counseling centers and digital-wellbeing programs. The findings support moving beyond the single question of how many hours students spend online to include timing, purpose, comparison processes, sleep-related effects, and emotional responses after engagement. Routine CES-D screening linked to clear referral pathways, digital-wellbeing workshops addressing self-awareness and healthy boundaries, peer and family gatekeeper training, and culturally responsive messaging that reduces stigma and normalizes help-seeking may all strengthen prevention. Because most participants were aged 18 to 20, spanning late adolescence and the transition to emerging adulthood, the evidence is also relevant to continuity in mental-health care as young people move from adolescent-oriented services to adult systems. Longitudinal, multi-institutional studies using richer measures of digital engagement — platform, activity type, timing, purpose, emotional response, and self-regulation — are needed to establish temporal ordering and test the proposed pathways. For now, the message is clear: the stressor students refuse to name may be the one most worth assessing.

Subject of Research: Mental Health

Article Title: A mixed-methods study of depressive-symptom severity among Thai university students and the underestimated role of social media

Article References: Boonkrong, P., Khunboonjun, K., Shakya, S., Yang, J., Vitoonmetha, N., & Chatchumni, M. (2026). A mixed-methods study of depressive-symptom severity among Thai university students and the underestimated role of social media. Frontiers in Child and Adolescent Psychiatry, 5, Article 1907727. https://doi.org/10.3389/frcha.2026.1907727

Image Credits: AI Generated

DOI: 10.3389/frcha.2026.1907727

Keywords: depression, social media, university students, Thailand, mental health, digital wellbeing, CES-D, mixed methods, mixed-methods, depressive-symptom, severity, Thai