Multiscale Framework for Bioclimatic Adaptation: Quantifying the Passive Performance of High-Mass Vernacular Heritage

Abstract

As global climate volatility intensifies, the built environment requires passive capacity to decouple habitability from external extremes. While vernacular architecture is a cited bioclimatic model, research often lacks long-term quantitative validation. This study addresses this gap through a multiscale framework applied to Montesinho Natural Park, Portugal. Integrating a typological survey with a one-year in situ monitoring campaign (2024–2025), the study utilises Python-based data processing to calculate statistical cross-correlations and benchmarks thermal resilience against the Portuguese Adaptive Comfort Model. Results substantiate a “Hierarchy of Filtration”: (1) Geological Scale: Location correlates statistically with lithological availability; (2) Settlement Scale: Topographical shielding suppresses the Diurnal Temperature Range (DTR) by 20.5%; (3) Envelope Scale: Traditional Stone-on-Earth assemblies exhibit a 16.5 h thermal lag, while vertical functional stratification dampens 47% of external annual temperature extremes. The study concludes that retrofitting must shift to “Balancing Inertia and Connectivity”. This approach mitigates the ‘maladaptation’ risks observed in modern lightweight interventions, providing an empirical template for passive thermal resilience applicable to resilient urban design in a warming climate.

1. Introduction

Vernacular architecture reflects generations of environmental adaptation, cultural tradition, and resource-conscious construction. The bioclimatic efficacy of vernacular architecture has long been acknowledged in foundational architectural theory. Rapoport [1] established that, while culture dictates form, extreme climates and available high-heat-capacity materials act as critical ‘modifying factors’ that compel traditional dwellings to adopt specific passive survival strategies. However, while this qualitative consensus exists historically, contemporary research often lacks the long-term, high-resolution quantitative validation required to translate these typologies into modern resilient design frameworks. In climatically demanding regions such as northern Portugal, traditional building practices offer valuable insights into passive thermal regulation, efficient material use, and spatial planning grounded in empirical local knowledge. However, as global climate volatility intensifies, bringing more frequent heatwaves and extreme storms, these traditional systems offer more than historical interest; they provide a blueprint for passive thermal resilience [2,3]. Modern resilient cities strive for what these vernacular settlements already possess: the passive capacity to decouple habitability from external climatic extremes without reliance on active energy grids.

Montesinho Natural Park (MNP), in Portugal’s Trás-os-Montes region, serves as a living repository of rural architectural wisdom. Its villages display varied construction systems shaped by lithological conditions, particularly granite and schist, and the climatic challenge of hot summers and severe winters. MNP was analysed not merely as heritage, but as a living laboratory of morphological and material resilience. By treating these settlements as compact “proto-cities”, we can extract tested genetic codes, morphological shielding, zoning, and inertia, which are scalable to contemporary urban design.

Prior studies have often addressed individual elements such as materials, building typologies, or village morphologies in isolation [4]. However, there is a lack of research exploring how these factors function as an integrated, synergetic bioclimatic system across multiple spatial scales. The present paper develops and proposes a multiscale interpretive framework. This methodology extends the analysis to the material and settlement scales, revealing how construction choices, envelope configurations, and spatial arrangements operate as reinforcing climate-responsive strategies.

This study addresses a key research gap: the need for systematic evidence of cross-scale bioclimatic adaptation within Mediterranean-Continental highland regions. While the specific architectural responses are inherently tied to the local lithology and climate of the Iberian Peninsula, the multiscale analytical framework developed here is designed to be adaptable to other European highland contexts. The MNP, with its diverse topography and geology, provides an ideal setting to test this framework. The primary objective is to explore how vernacular architecture, in a specific microclimatic environment, e.g., Portugal, achieves environmental coherence through:

• Geological Scale: Statistical dependencies between village location and lithological availability, interpreted through a lens of thermal persistence.
• Envelope & Material Scale: Typological classification of wall and roof systems, interpreted through passive performance criteria and functional stratification.
• Settlement Scale: Morphological patterns, including clustering and slope-responsive layouts, read as adaptive spatial strategies.

The paper adopts a mixed-methods approach, combining field-based typological analysis with a preliminary quantitative diagnostic. While the multiscale framework remains the central qualitative contribution, it is supported by an in situ long-term hygrothermal monitoring validation to check the physical efficacy of the identified strategies. Critically, this methodology is presented as an expandable framework, intended to be a structured basis for future high-fidelity simulations or participatory retrofit strategies. The findings provide localised insights for heritage-sensitive design, suggesting that this vernacular ecological intelligence can inform contemporary rural resilience efforts within similar Mediterranean-Continental contexts.

2. Theoretical Framework: The Multiscale Deficit

Vernacular architecture is increasingly recognised as a multiscale system of environmental adaptation shaped by local climate, cultural tradition, and resource-responsive construction [1,5,6,7]. Classic works by Fathy (1986) [8], Olgyay (1963) [9], and Watson and Labs (1983) [10] laid the groundwork for understanding how traditional buildings respond passively to environmental stimuli through material use, orientation, and spatial planning. More recent studies have expanded this understanding across global contexts, emphasising the critical need for systemic, scale-sensitive analysis. Recent theoretical frameworks underscore that vernacular bioclimatic evolution must be evaluated through a multi-scale lens [11], while field studies in the Mediterranean have successfully demonstrated how building-level performance is inextricably linked to urban-scale morphology [12]. For example, Rijal (2021) [13] demonstrates that, in extreme high-altitude climates, building adaptation via thick earthen mass is essential to maintain indoor temperatures significantly higher than the outdoor ambient.

Drawing on foundational work by Oliver (2006) [6], which defines vernacular architecture as a complex interplay of cultural expression and environmental adaptation, this study extends that systemic perspective into a structured multiscale framework. Likewise, the Atlas of Vernacular Architecture [7] emphasises the global role of resource-based construction systems and spatial patterns, underpinning the methodology of scale-sensitive typological analysis. This is further formalised by the VerSus project [14], which provides a comprehensive framework for identifying sustainable vernacular strategies across three levels: “Settlement”, “Building”, and “Detail”, specifically highlighting the “environmental intelligence” found in traditional European regions.

At the material scale, the research highlights how locally sourced materials enhance passive thermal and hygrothermal regulation, such as rammed-earth structures in hot-arid climates [8], which optimise thermal mass for daily temperature cycles; stone-timber farmhouses in Eastern Tibet [15], combining thick masonry walls with ventilated skylights and wooden insulation layers; and high-altitude Ladakhi dwellings with composite wall-roof systems that manage both heat retention and moisture [16]. These examples affirm the role of material logic in environmental adaptation. In high-altitude Nepal, Rijal (2021) [13] identified that 450 mm thick dry brick walls and mud roofs act as primary passive heating elements, effectively storing heat for nocturnal release.

At the building scale, strategies like thick walls, shaded courtyards, and compact forms have been documented in Andean, Mediterranean, and North African contexts [9,17]. This includes functional zoning where vertical stratification—using ground floors for storage or livestock—creates a thermal buffer for inhabited spaces.

At the settlement scale, traditional vernacular communities in Mediterranean islands such as Cyprus and Greece, as well as in Nepal’s hill towns, demonstrate clustering, slope orientation, and solar alignment strategies that reflect bioclimatic design principles [18,19,20]. These spatial arrangements effectively manage solar gain and shade to mitigate heat and cold in mountainous terrain, consistent with foundational bioclimatic theory [9,10]. Rijal (2021) [13] specifically notes that courtyard houses in Himalayan regions are attached to one another to protect residents from extreme winds, a form of collective thermal shielding.

Despite growing interest in vernacular sustainability, many studies remain siloed within a single scale, focusing either on construction typologies or spatial layouts in isolation. This fragmentation is critiqued by Vellinga (2013) [5] and by Correia, Dipasquale, and Mecca (2014) [14], who call for integrative approaches across material, building, and settlement scales. Examples of such holistic frameworks remain limited in practice, particularly in Middle Eastern reviews (e.g., [21]). For example, while Martins and Nóvoa (2015) [22] provide detailed descriptions of granite masonry in northern Portugal, they rarely link this to broader village morphology or climatic planning strategies.

Broader scholarship has emphasised the importance of cross-scale coherence, where material, envelope, and settlement-level decisions are viewed as part of an integrated bioclimatic system [16,17]. However, in rural European contexts, particularly within the Iberian Peninsula, such frameworks are still scarce. The MNP in north-eastern Portugal remains largely underexplored in this regard. While prior research has documented local materials and climate [23,24], few studies adopt a multiscale, field-based analytical model.

This study addresses that gap by combining observational, typological, and statistical methods to examine bioclimatic adaptation across three spatial scales. By incorporating a preliminary quantitative diagnostic, the research moves beyond descriptive approaches toward performance-oriented interpretations of the systemic environmental reasoning embedded in traditional architecture.

3. Materials and Methods

This study adopts an exploratory, multiscale field-based methodology to investigate how vernacular architecture in the Montesinho Natural Park (MNP) responds to its severe, high-altitude climate (Figure 1), reflecting bioclimatic adaptation across three nested spatial scales: material selection, envelope configuration, and settlement morphology. The approach builds on previous studies, one focused on regional digital surveys [25] and another on building-scale typological analysis [4], by integrating field observation, typological interpretation, frequency analysis, and spatial reading across selected villages. This framework is further supported by a preliminary quantitative diagnostic to benchmark physical performance against the qualitative findings.

3.1. Research Strategy

According to the Köppen–Geiger climate classification, the MNP is characterised primarily by a warm-summer Mediterranean climate (Csb) [27,28,29], featuring severe, prolonged winters and hot, dry summers. To define representative case studies, a multiscale characterisation of the region was performed:

(a)

Geological Scale (Regional Survey): A preliminary inspection of 1620 buildings across 13 villages was conducted to map the dominance of granite and schist in wall, roof, and opening materials.

(b)

Envelope & Material Scale (Typology): Typological analysis focused on a subset of buildings with intact vernacular characteristics, specifically examining building external composition.

(c)

Settlement Scale (Bioclimatic Configuration): Spatial analysis utilised aerial imagery, topographic data, and detailed 3D photogrammetric models to evaluate the village clusters’ responses to environmental exposure. From this analysis, the village of Pinheiro Novo was purposefully selected for its atypical West-facing slope to challenge standard assumptions about passive solar orientation and topographical shielding, serving as a strategic counterpoint to the conventional South-facing clusters.

3.2. Geomorphological Context (Geological Scale) and Case Study Selection

3.2.1. Lithological Domains

In vernacular architecture, the envelope is central to thermal performance, especially in climatically challenging contexts like MNP. Traditional walls, roofs, and floors rely on material thickness, porosity, and layering to regulate heat passively, despite lacking modern insulation. These strategies reflect both environmental adaptation and local resource constraints.

Field observations across multiple villages within MNP reveal four principal external wall typologies, each reflecting distinct traditional construction logics and material strategies shaped by geological availability and vernacular knowledge. These are classified as: (1) Dry Masonry, (2) Ashlar Masonry, (3) Irregular stone Masonry with mortar, and (4) Tabique (timber frame with infill). Their prevalence and performance characteristics are not uniform but vary in response to settlement patterns, construction period, and material sourcing.

Regional studies offer useful proxies, for instance, Pereira (2016) [30] analysed the thermal behaviour of granite and schist walls in Trás-os-Montes, while Barroso et al. (2020) [31] documented the physical properties of similar materials in dry-stone construction. Given the geological and typological continuity, these references provide a reasonable basis for interpreting the thermal tendencies of MNP’s masonry.

(a)

Type 1: Dry Masonry

Dry masonry walls, composed of uncut granite or schist stones laid without mortar, are prevalent in older structures in villages such as Pinheiro Novo and Guadramil. These walls typically exceed 60 cm in thickness, providing high thermal inertia that passively buffers indoor conditions against daily temperature fluctuations. However, the absence of mortar results in low airtightness, making such constructions vulnerable to winter infiltration and discomfort.

(b)

Type 2: Ashlar Masonry

Ashlar masonry employs carefully cut stones with tight-fitting joints, typically found in more recent or socially prominent constructions. This typology improves structural precision and reduces voids between stones, offering greater durability and reduced thermal bridging compared to dry masonry. While less common, it is particularly evident in transitional buildings or those with symbolic or representational functions. The use of this technique aligns with broader patterns of stone refinement documented in regional heritage contexts [32,33].

(c)

Type 3: Mortared Masonry

Mortared stone walls are the most common typology across the park, incorporating lime- or earth-based binders depending on the village and construction period. These walls typically use rubble or semi-cut stones with mortar applied inconsistently, resulting in variable thermal and hygric performance. According to national technical codes of traditional envelopes, such walls generally exhibit moderate thermal resistance, with performance highly dependent on joint continuity and workmanship quality [22,34].

(d)

Type 4: Tabique

Though less frequent, tabique walls, light timber frames filled with earth, stone, or lime plaster, are present in some auxiliary or infill structures within MNP. This technique has been historically widespread in Trás-os-Montes and Alto Douro and is recognised as part of the regional bioconstructive heritage [35,36]. While their thermal mass is lower than stone assemblies, tabique walls offer flexibility, ease of repair, and contribute to internal moisture regulation. Recent heritage manuals identify their cultural and constructive value, especially in mixed-wall systems or secondary building elements [34,35].

Although not specific to MNP, the masonry typologies documented in LNEC ITE 54 [34], developed for pre-1950s Portuguese buildings, offer a useful reference point. Some descriptions align closely with vernacular wall constructions observed in MNP and are therefore used here to support typological categorisation. A synthesis of the most common masonry types identified in the region, providing brief descriptions and representative images, is shown in

Table 1. Masonry types and descriptions, source: (Santos & Rodrigues, 2009) [34].

Example	Masonry Category	Description
Rio de Onor	Simple Ordinary Masonry walls (Schist/Granite)	Materials: Stones of varying sizes (rough or partially refined); occasionally integrated with brick fragments. Joints: Mortar proportion is typically substantial, often surpassing 25%. Finishes: Plaster (air lime/sand) typically ranges from 70–120 mm [37]; other coverings (e.g., ceramics or slate) may also be observed. Thickness: Generally ranges between 0.4 m and 0.9 m.
Pinheiro Novo	Simple Dry Masonry walls (Schist/Granite)	Materials: Coarse schist or granite (angular or rolled) with irregular shapes and varying dimensions. Joints: Typically laid without mortar on the most regular face; voids are commonly packed with earth, straw, or stone chips to reduce air infiltration. Finishes: Usually plastered on the interior face only. Features: Generally characterised by high wall thickness and small fenestration openings.
Pinheiro Novo	Simple ashlar stonework(Exterior)	Materials: Stone in which all faces are generally prepared, worked, or trimmed. Joints: Usually thin and filled with mortar; however, stones may be dry-stacked (overlapped without mortar) if surfaces are sufficiently even.

. While the images represent only selected examples within broader categories, they help clarify construction traits relevant to passive thermal behaviour.

3.2.2. Lithological and Typological Verification

In order to quantify this typological and lithological distribution, a comprehensive visual census of 1620 vernacular buildings across 13 villages within the MNP was conducted. This protocol involved cross-referencing Geographic Information System (GIS) mapping with extensive in situ field surveys, cataloguing individual structures based on their primary load-bearing material to establish the concrete statistical baseline.

To ensure the monitored case studies represent the dominant vernacular forms rather than anomalies, the selection was justified using the statistical dependency protocol established in [4]. This analysis quantifies the relationship between local lithology and construction techniques using standardised frequency residuals (N(0,1)). By cross-tabulating village locations with observed building materials, residuals were calculated for each village–material pair. Values exceeding the standard normal critical threshold (1.644) indicate a statistically significant positive dependency, demonstrating that a material’s local prevalence represents a deliberate vernacular adaptation rather than random chance.

The results (Figure 2) reveal a statistically significant dependency between the granite bedrock and the masonry typology in the selected villages. Pinheiro Novo exhibits the highest dependency value for granite wall construction (10.37), followed by Montesinho (9.01), albeit these areas are adjacent to both materials (granite and schist), suggesting that superior mechanical properties [31] may outweigh mere proximity. Northern villages, namely Montesinho (76.1%), Pinheiro Novo (43.2%), and Rio de Onor (42.5%), show a higher prevalence of slate roofs.

The distribution of these geological zones across the MNP is illustrated in Figure 3 (which also highlights Pinheiro Novo and Montesinho villages, the eventual sites selected for the micro-scale thermal monitoring detailed in Section 3.2.3). This aligns with the distribution of slate-bearing schists, identified in the more detailed 1:200,000 geological mapping (generally comparable with legend code “O_a” in Figure 3), which offered locally available and workable stone for roofing, even if not always of commercial-grade quality. Furthermore, the analysis is highly consistent with high dependency values for slate roofing (11.93) and traditional timber fenestration (3.64) in the Montesinho village.

3.2.3. Case Study Selection and Characterisation

In order to evaluate the multiscale performance framework, three vernacular archetypes were selected within the monitored villages (Figure 4 &

Table 2. Specific Architectural, Spatial, and Operational Characteristics of the Monitored Case Studies.

Feature Category	Case 1 (The Vernacular Baseline)	Case 2 (The Modern Reconstruction)	Case 3 (The Sensitive Rehabilitation)
Intervention Logic	Non-Intervention/Original structure	Modern “Linear” Reconstruction	Adaptive reuse/Residential conversion
Building Adjacency	Terraced (Clustered morphology sharing lateral stone party walls)	Terraced (Inserted between existing granite party walls)	Semi-detached (High exposure: 3 external facades, sharing only the SW party wall)
Monitored Zone Volume	13.5 m3 (1st-floor bedroom)	38.5 m3 (1st-floor rear bedroom)	50.7 m3 (1st-floor rear bedroom with vaulted sloped ceiling)
Spatial Zoning & Exposure of Monitored Zone	“Sandwiched”: Primarily buffered by adjacent internal rooms	Corner Exposure: Two external walls (Highly glazed SE façade + completely opaque NE façade)	Roof & Façade Exposure: Volume extends directly to the pitched roof (lacking the attic buffer present in the building’s centre), alongside an exposed Southeast external wall
Façade Window-to-Wall Ratio (WWR)	~0% (Internal/indirect light only)	High (42%) glazed façade facing SE	Low (4%) façade facing SE
Glazing & Frame Specification	N/A (No external window in monitored zone)	Double-glazed clear glass with timber frame	Double-glazed clear glass with timber frame
External Wall Structure	Variable thickness: 60–80 cm double-leaf dry granite masonry	Facades: Reinforced concrete frame infilled with thermal clay blocks. Party Walls: Retained original granite masonry (structurally integrated with new concrete frame).	Original granite masonry (Retained shell)
Roof Assembly	Traditional slate and ceramic tiles on uninsulated timber frame (Highly permeable). Includes a naturally ventilated, non-habitable attic buffer	Slate tiles over insulated (XPS) concrete slab with standard ceiling finish. No attic buffer (unventilated structure directly coupled to monitored bedroom)	Slate tiles over insulated (EPS) concrete slab with recreated wooden ceiling. No attic buffer over the monitored zone (unventilated vaulted ceiling)
Insulation Strategy	None (Relies entirely on “lithological inertia”)	Internal lining: XPS insulation + laminated plasterboard	Internal lining: Thermal plaster with EPS aggregate
Heating System	Indirect Radiant (Borrowed heat from adjacent kitchen wood stove)	Mobile Convective (Electric unit—on/off demand)	Centralised Hydronic (Radiators—intermittent use)
Occupancy Profile	Continuous (Full-time elderly resident)	Intermittent (approx. 2 weeks per month)	Intermittent (Secondary dwelling/Short stay)
Material Circularity Logic	Indigenous (100% Circular/Zero embodied carbon)	Material Discontinuity (Linear substitution/Industrial concrete logic)	Structural Circularity (Retains heavy shell; compromised by synthetic layer)
Thermal Goal	Passive Thermal Resilience	Instant Convective Comfort (Low Inertia)	Normative Habitability

), representing distinct stages of the region’s building stock evolution.

(a)

Case 1 (The Traditional Baseline): Located in Pinheiro Novo with a Northeast primary orientation, this building represents the original “Stone-on-Earth” typology. It features 60–80 cm uninsulated dry-stone granite walls and single-glazed wood frames, occupied full-time by an elderly resident. The monitoring analysis focuses on the bedroom (13.51 m³), which functions as a “sandwiched” thermal zone: it is structurally separated from the primary heat source (kitchen firewood stove at the back) by a 12 cm brick partition and buffered by a study room at the front.

(b)

Case 2 (The Modern Reconstruction): Situated in Montesinho village, this case employs a hybrid structure of reinforced concrete, incorporating internal insulation and double glazing. While the building has a general Northwest orientation, the analysis focuses on a rear bedroom (38.45 m³) with a Southeast-oriented façade. This specific zone features a high façade window-to-wall ratio (WWR) of 0.42, serving as a proxy for analysing high solar gain admission and night-time ventilation potential in a thermally tight yet structurally heterogeneous system heated by a mobile electric unit.

(c)

Case 3 (The Rehabilitated Secondary Dwelling): Also located in Montesinho (Northwest orientation), this case represents an internally insulated rehabilitation, retaining the granite appearance but integrating a continuous layer of internal insulation and high-performance frames. It is the only case equipped with a centralised hydronic heating system. The assessment explicitly monitors a specific vaulted 1st-floor bedroom (50.70 m³) with an external wall facing Southeast. This zone functions without proper cross-ventilation, isolating the interaction between internal insulation and solar gain to represent the modern standard for intermittently occupied rural renovations.

3.3. Data Acquisition and Monitoring

The empirical foundation of this study relies on in situ monitoring conducted using Testo 174H data loggers (Testo SE & Co. KGaA, Titisee-Neustadt, Germany) recording at 30 min intervals over a continuous annual cycle (February 2024–February 2025). To ensure statistical comparability across typologies and eliminate installation-phase thermal noise, a standardised 365-day reference window of 8760 h (9 February 2024, 00:00 to 9 February 2025, 00:00) was enforced for all statistical calculations, unless otherwise stated (e.g., the 362-day synchronised window utilised for Case 3 comparative metrics in Section 4.1.2). These devices are equipped with NTC temperature and capacitive humidity sensors that comply with ISO 7726 Class C accuracy standards [38]. Specifically, they offer a resolution of 0.1 °C and 0.1% RH, with a certified operating accuracy of ±0.5 °C and ±3% RH (at 25 °C). To ensure data integrity and prevent measurement anomalies, all devices were strategically positioned away from direct solar radiation and active heat sources.

Due to the logistical constraints of monitoring occupied dwellings, specifically the need to prevent sensor tampering or obstruction, installation heights varied by typology. In Case 1, the logger could be safely placed at bedside level (0.7 m) to directly capture the occupied thermal plane. However, in Cases 2 and 3, sensors had to be installed at elevated positions (2.0–3.0 m) to keep them out of reach and prevent interference by intermittent occupants. Due to the physics of thermal stratification, this elevated placement in Cases 2 and 3 introduces a deliberate methodological bias: winter data represent a conservative estimate of cold stress, as the actual occupied zone is likely colder than recorded; conversely, summer data capture a maximum heat accumulation scenario, providing a robust metric for evaluating the risk of overheating in insulated volumes.

To isolate the passive performance of the vernacular systems from occupancy variables, the sensor deployment followed a dual-scale strategy:

(a)

Envelope & Material Scale: Internal sensors were installed in the central bedrooms of all three cases (as defined in Section 3.2.3). This single dataset serves a dual analytical purpose:

• Inertia Metrics: To derive the Dynamic Thermal Lag (τ) and amplitude damping metrics by isolating strictly ‘passive windows’ (periods where active heating is entirely inactive).
• Long-Term Habitability: To generate the annual Adaptive Comfort profiles (discussed in Section 4.3) by recording continuous temperatures across all seasons.

(b)

Settlement Scale: To quantify microclimatic moderation, the external ambient sensors of Pinheiro Novo (Case 1) and Montesinho (Case 3) were cross-referenced. This comparison isolates the Diurnal Temperature Range (DTR) suppression resulting from the specific topographical shielding and morphological clustering of each village, independent of the building envelope.

3.4. Data Processing and Performance Assessment

To evaluate both the intrinsic physics and the human-centric habitability, the acquired datasets were filtered and processed according to two distinct protocols. The complete analytical workflow, from raw data acquisition through temporal filtering to the extraction of the multiscale performance metrics, is summarised in Figure 5.

3.4.1. Passive Metrics (Inertia Analysis)

The analysis of thermal performance was executed through a custom-developed Python (v3.11) analytical workflow (Python Software Foundation, Wilmington, DE, USA). Prior to statistical calculation, the raw environmental datasets underwent a rigorous data-handling protocol (as mapped in Figure 5). Time-series arrays from multiple sensors were synchronised and verified for a uniform 30 min frequency. Minor sensor dropouts were resolved utilising linear interpolation to maintain the strict signal continuity required for algorithmic processing. Rather than a purely descriptive approach, the script implemented a Cross-Correlation Algorithm to provide mathematically rigorous evidence of thermal behaviour. Four primary metrics were utilised:

(a)

Thermal Stability Index (SI): The standard deviation of indoor temperatures (σ_Ti), quantifying the passive persistence and resistance to external volatility.

$$\mathrm{S}\mathrm{I}={\sigma }_{T_i}=\sqrt{{\displaystyle \frac{1}{N-1}}{\sum }_{i=1}^N{\left(T_i-\overline{T}\right)}^2}$$

(1)

(b)

Diurnal Temperature Range (DTR) Moderation: A comparison of daily external temperature amplitudes within village clusters to quantify the microclimatic “shielding” effect of the morphology. The daily DTR is calculated as:

DTR_daily = T_max,daily−T_min,daily

(2)

(a)

Thermal Amplitude Damping: The capacity of the envelope to mitigate external thermal loads, quantified via two complementary indices:

• Decrement Factor (f): A dimensionless ratio of the indoor to outdoor temperature amplitudes during the 168 h reference week.

  $$f={\displaystyle \frac{T_{in,max}-T_{in,min}}{T_{out,max}-T_{out,min}}}$$

  (3)
• Absolute Damping (ΔA): The absolute temperature suppression achieved at the peak environmental load, providing an intuitive metric for passive cooling efficacy.

$$\Delta A=T_{out,max}-T_{in,max}$$

(4)

(b)

Dynamic Thermal Lag (τ): The temporal offset (phase shift) between standardised outdoor and indoor thermal signals. This is determined by calculating the cross-correlation of the Z-score standardised arrays to identify the time delay that yields the maximum correlation coefficient.

$$\tau =\arg \underset{\Delta t}{\max }\left(\sum {(T}_{out}\left(t\right)\cdot T_{in}\left(t-\Delta t\right)\right)$$

(5)

A 168 h reference week (15–21 July 2024) was algorithmically extracted by identifying the period of peak external thermal amplitude. This maximum environmental load provides the clear rhythmic driving signals required for cross-correlation while guaranteeing a true ‘passive window’ (where active heating is entirely inactive), thereby capturing the pure physical performance of the envelope.

3.4.2. Adaptive Thermal Comfort Assessment

To evaluate the combined effectiveness of the multiscale strategies on habitability, indoor operative temperatures were benchmarked against the Portuguese Adaptive Comfort Model [39]. Because international standards (such as ASHRAE 55 [40] or EN 15251 [41]) do not fully capture regional tolerances, this model adapts those international standards specifically to the Portuguese context. It was selected because it accounts for the typical Mediterranean climate, local ways of living, and how national buildings are conventionally designed and used, acknowledging that occupants of naturally conditioned buildings tolerate broader temperature ranges than those in mechanically climatised spaces. This approach determines the percentage of time indoor temperatures fall within the acceptability boundaries defined by the running mean outdoor temperature of the preceding 7 days (Θ_rm).

(a)

The Operative Temperature Proxy: Following established field monitoring practices in vernacular contexts [42], indoor air temperature (Θ_a) serves as a robust proxy for operative temperature (Θ_op) across all three typologies. While the use of globe thermometers would provide higher precision regarding radiant asymmetry, this approximation is physically justified for the specific typologies analysed:

• Case 1 (Passive/Indirect): The absence of direct heating emitters (warmth relies on indirect transfer from the adjacent kitchen) combined with high envelope inertia ensures uniform surface temperatures.
• Case 2 (Convective/Decoupled): The use of mobile electric heaters provides primarily convective heat, directly conditioning the air volume. Crucially, while the building utilises a reinforced concrete structural frame, its envelope system—composed of thermal clay blocks, XPS insulation, and an internal drywall lining (plasterboard)—completely decouples the indoor space from any structural thermal mass. This creates a low-inertia inner surface that equilibrates rapidly with the air temperature, preventing the significant radiant asymmetry that would otherwise occur with high-intensity spot heating.
• Case 3 (Intermittent/Insulated): Although occupancy is seasonal, the internal insulating plaster similarly decouples the indoor space from the cold granite mass. This creates a responsive inner lining that tracks air temperature fluctuations during heating cycles, validating the proxy even under intermittent conditions.

(b)

The Adaptive Limits: The model applies distinct algorithms to account for the differing thermal expectations of occupants:

• Case 1 (Naturally Ventilated): Reflecting the specific adaptation to free-running vernacular conditions, the comfort temperature (Θ_c) follows Matias’s calibrated algorithm (slope 0.43). The applied ±3.0 °C bandwidth corresponds to Category II (Normal Expectation) as defined in EN 15251 [41] (90% acceptability):

  Θ_c = 0.43⋅Θ_rm + 15.6 (±3.0 °C)

  (6)
• Cases 2 and 3 (Intermittently Heated): To account for the “modern expectation” of occupants relying on intermittent active heating, the model adopts the standard ASHRAE 55 adaptive coefficient [40,43] which has been validated for the Portuguese context in the TP165 guidelines [39]. A wider ±3.5 °C bandwidth is applied (80% acceptability):

Θ_c = 0.31⋅Θ_rm + 17.8 (±3.5 °C)

(7)

4. Results

This chapter evaluates the environmental performance of the vernacular case studies, tracing the hierarchy of passive resistance from the urban tissue down to the occupant’s experience. The analysis moves beyond basic architectural description to systematically identify the specific bioclimatic strategies embedded in the built form and quantitatively verifies their efficacy through long-term in situ monitoring data.

The results are structured across three interacting scales:

(a)

Settlement Scale (Section 4.1): Examining how morphological clustering and orientation create a “microclimatic shield” that tempers external exposure and suppresses the Diurnal Temperature Range (DTR).

(b)

Envelope & Material Scale (Section 4.2): Quantifying the thermal inertia, phase shift, and amplitude damping of the traditional granite construction system.

(c)

Human Scale (Section 4.3): Assessing the final impact on seasonal habitability using the Portuguese Adaptive Comfort Model.

4.1. Settlement Scale: Spatial Organisation and Climatic Strategies

4.1.1. Aerodynamic Strategies: Shielding (Compact) vs. Channelling (Linear)

Based on our direct field observations, topographical mapping, and drone photogrammetry—correlated with local wind profiles generated via Climate Consultant using Bragança station data—the settlement morphology in the MNP reveals two distinct aerodynamic responses to the seasonal wind regime (Figure 6).

Wind Exposure as a Critical Challenge

Wind exposure presents a critical climatic challenge in the region. Historical winter data from Bragança (1971–2000) indicate average wind speeds of 9–11 km/h, with extreme gusts reaching 115–140 km/h [44]—levels classified as “violent storm” to “hurricane” on the Beaufort Wind Scale [45]. Even modest wind increases (e.g., <0.5 m/s) can significantly alter thermal sensation, with studies indicating that air movement can lower the effective temperature by ~1 °C [46,47].

(a)

The Compact Shielding Strategy (Pinheiro Novo):

The high-altitude villages (830–1030 m) typically adopt a compact cluster typology, as exemplified by Pinheiro Novo (Figure 7). This aggregation is a direct defence against the high-intensity winter winds from the North and Southeast.

• Siting Logic (Water vs. Sun): Pinheiro Novo is sited on a West-facing slope (830–850 m) to maintain proximity to the Rabaçal river tributaries and fertile agricultural terraces. This socio-economic priority forces a critical trade-off: the location naturally lacks optimal South-facing solar exposure.
• Aerodynamic Defence and Building Adjacency: To survive this exposed positioning, the settlement adopts a dense, irregular morphology heavily dictated by local geology and geomorphology. As shown in Figure 7, the village is strategically sited on a relatively gentle geological terrace nestled between higher protective mountain ridges to the Northeast and a steep river valley descent to the West. Because such moderate terrain is scarce within this rugged geological landscape and historically preserved for agricultural use, the buildable footprint is tightly constrained. This, in part, forces a high-density agglomeration where individual dwellings are rarely isolated. Instead, buildings share multiple party walls. This adjacency is a fundamental passive strategy: by clustering together, the structures drastically reduce their exposed external envelope area (lowering the surface-area-to-volume ratio), which directly minimises winter convective heat loss and provides collective thermal mass buffering. This compact shielding compensates for the lack of solar gain by minimising the envelope’s exposure to the severe North/Southeast winter winds, prioritising heat retention over passive solar potential. While this aerodynamic interpretation is based on qualitative spatial and topographical analysis, its physical efficacy is quantitatively corroborated by the microclimatic sensor data (DTR suppression) presented in Section 4.1.2.
• The Road Network as a Climatic Filter: Furthermore, the circulation patterns are not merely logistical; they act as an extension of the climatic shield. The road network organically traces the geological contour lines of the terraces, resulting in narrow, tortuous pathways. This irregular street geometry deliberately breaks laminar airflow, preventing the cold winter winds from “tunnelling” through the village, while the narrow street profiles ensure mutual shading between adjacent structures during peak summer afternoons.

(b)

The Linear Channelling Strategy (Rio de Onor):

In contrast, the valley-based Linear Typology (Figure 8) of Rio de Onor reflects an opposing logic driven by ventilation and topographic adaptation.

• Winter Alignment and Circulation: The village aligns North–South along river banks. Unlike the compact clusters that clump together on steep terrain for warmth, this linear form strictly follows the valley’s geological contours to preserve the fertile agricultural land on the river margin. The road network functions as a central circulation spine. While this orientation exposes the settlement to the cold N/S/SE winds, the linear street grid allows winds to pass through the settlement corridor rather than accumulating static pressure against the building facades. Adjacent buildings are attached sequentially along this road, protecting their lateral walls while leaving the primary front and rear facades exposed for interaction with the street and river.
• Summer Ventilation (NW/W Dominance): During the cooling season, the prevailing Northwest and West winds strike the linear cluster transversely (almost perpendicular to the long facades). Unlike the compact clusters, which block wind, this exposure creates a pressure differential across the shallow building depth, facilitating effective cross-ventilation when occupants open fenestration on opposing front and rear facades.
• Solar Exposure (East–West): Consequently, the primary facades are oriented East and West, deviating from the theoretical South-facing ideal. Instead of steady Southern gain, the buildings accept dynamic solar exposure: receiving direct morning (East) and afternoon (West) sunlight. This morphological outcome parallels the historic Solskifte (“solar village”) typologies of Northern Europe, identified by Rapoport [1] and further documented in medieval settlement studies [48] as a vernacular strategy to ensure equitable solar distribution along a North–South street axis. In Rio de Onor, this linear East–West exposure ensures that, even in deep winter, the facades capture solar access during the limited hours when the sun clears the steep valley ridges. Recent computational simulations [49] indicate that such longitudinal street orientations in mountainous terrain are highly effective at balancing this dynamic insolation with the need for longitudinal ventilation.

This typological divergence suggests that MNP vernacular architecture is not a singular formula, but a site-specific aerodynamic calibration. Where the valley topography provides natural shelter (Rio de Onor), the morphology is free to adopt a linear channelling logic, aligning with the axis to facilitate ventilation. Conversely, where the highland terrain creates severe exposure (Pinheiro Novo), the morphology is forced into a compact shielding logic, clustering together to resist turbulent wind loads. These findings have direct implications for resilient urban planning, demonstrating that morphological density is a critical asset for aerodynamic defence in volatile climates.

4.1.2. Quantitative Validation: Microclimatic Stability and Topographical Calibration

Settlement-scale validation compares the external ambient conditions of the compact clusters in Pinheiro Novo and Montesinho to characterise the efficacy of topographical positioning. By processing high-resolution sensor data (T4 and M2) over a standardised annual cycle (February 2024–February 2025), the research identifies how cluster morphology serves as the primary environmental filter, dampening the raw volatility of the mountain climate before it reaches the building envelope. Due to some sensor battery depletion in Case 3, a synchronised 362-day monitoring period (9 February 2024–5 February 2025) was utilised to ensure cross-case statistical comparability.

The results, summarised in (

Table 3. Village-Scale Diurnal Temperature Range (DTR) Comparison.

Settlement Cluster	Altitude	Orientation/Cluster	Avg. Annual DTR [°C]	Peak Daily Range [°C]
Pinheiro Novo (Case 1)	~840 m	West-Facing, Compact	4.68	15.90
Montesinho (Case 3) *	~1010 m	South-Facing, Compact	5.89	17.60

Note: * 2 February 2025 12:00:00 to 5 February 2025 00:00:00. External temperatures extracted from Case 2.

), establish a distinct “stabilisation signature” for each settlement type, demonstrating that vernacular location is a calculated trade-off between energy harvesting and environmental protection.

(a)

Geomorphological Shielding

The data for Pinheiro Novo are highly consistent with a high degree of topographical protection. Despite its mid-altitude location, the village maintains a remarkably stabilised microclimate with a DTR of 4.68 °C. This stability is attributed to its West-facing, mid-slope positioning, where the surrounding mountain mass to the North and Southeast acts as a kinetic and thermal shield. This topographical sheltering mirrors strategies observed in other northern Portuguese vernacular contexts, such as the Douro region, where valley implantation is utilised to mitigate extreme thermal amplitudes [50]. By dampening the daily thermal cycle, the topography reduces the mechanical stress on the stone masonry. The peak range of 15.90 °C (recorded in June 2024) indicates that, while the settlement is geomorphologically optimised for winter sheltering, it maintains a resilient baseline that prevents the “over-volatility” common in higher, more exposed peaks.

(b)

Solar Harvesting and Volatility

Montesinho exhibits a more dynamic microclimate (Avg. Annual DTR = 5.89 °C) compared to the shielded topography, leeward slope location of Pinheiro Novo (4.68 °C), confirming that the latter achieves a 20.5% suppression in diurnal variability. This is a direct consequence of its South-facing “amphitheatre” morphology at high altitude (≈1010 m), a strategic positioning that prioritises solar harvesting for thermal gain in the “Cold Land” (Terra Fria).

While this orientation increases radiative exposure, the compact, agglomerated cluster morphology provides critical aerodynamic protection. The building-to-building proximity functions as a buffer against prevailing winter Northeastern winds, breaking laminar flows and reducing convective heat loss. The effectiveness of this mutual shielding is evidenced by the restricted peak DTR of 17.60 °C, which, while higher than the sheltered Pinheiro Novo, remains significantly lower than regional open-field averages, which can exceed 20 °C during peak summer months in the high-altitude plateaus of Trás-os-Montes [26,44].

Thus, the Montesinho cluster shows a calibrated trade-off: accepting higher solar volatility to ensure thermal survival while utilising morphological density to mitigate high-altitude wind stress. In the context of climate change, where diurnal extremes are predicted to intensify, this capacity to passively suppress volatility (reducing DTR peaks) serves as a critical benchmark for ‘Urban Resilience’, demonstrating that morphological compactness is a functional necessity for climatic stability.

4.2. Envelope & Material Scale: Envelope Performance and Spatial Layering

4.2.1. Thermal Inertia: Absolute Damping and Time Lag

The material-scale efficacy was evaluated using a Z-score standardised cross-correlation algorithm by comparing the entire 168 h wave pattern (15–21 July) to determine the dynamic Thermal Lag (τ), Decrement Factor (f), and Absolute Damping (ΔA) (

Table 4. Dynamic Thermal Performance Benchmarking (Summer Period: 15–21 July 2024).

Case Study	Typology	Wall Envelope	Thermal Time Lag (τ) [h]	Decrement Factor (f)	Absolute Damping ΔA [°C]
Case 1	Traditional	Granite Masonry (Not insulated)	16.5	0.259	3.80
Case 2	Modern	Reinforced concrete frame infilled with thermal clay blocks (Internal insulation)	0.0	0.076	2.10
Case 3	Rehabilitation	Granite Masonry (Internal insulation)	0.0	0.36	−3.10 *

* Note: The negative Absolute Damping (−3.10 °C) in Case 3 indicates an indoor environment that exceeds external peak temperatures. This may be attributed to the increased airtightness and internal insulation of the renovated envelope, which creates a strong decoupling from the external environment but traps internal heat, eliminating the convective heat dissipation (night cooling) characteristic of the more permeable vernacular typology.

). This benchmarking contrasts the passive persistence of traditional granite masonry (Case 1) against the thermal volatility of modern reinforced concrete reconstruction (Case 2) and the unintended consequences of internally insulated rehabilitation (Case 3).

The analysis (Table 4) identifies three distinct material behaviours:

(a)

Traditional Case 1: The superior performance of the Case 1 bedroom is a result of spatial-material synergy. The sensor data reveal that the centrally located bedroom achieves exceptional stability, exhibiting a 16.5 h thermal lag and an Absolute Damping (ΔA) of 3.80 °C. By placing the sleeping quarters in the middle of the building, decoupled from the external granite walls, the traditional typology uses the surrounding rooms as an architectural buffer. This spatial zoning works in tandem with the massive 60–80 cm granite envelope, ensuring the core remains completely immune to daytime peaks by shifting the thermal load entirely into the following nocturnal cycle.

(b)

Rehabilitation Case 3: Case 3 fundamentally fails to replicate the passive persistence of the vernacular envelope, exhibiting a near-instantaneous 0.0 h thermal lag and a negative Absolute Damping (−3.10 °C). While the addition of modern internal insulation successfully ‘stops’ external heat conduction, it concurrently triggers a ‘Thermos Effect’. The internal insulation critically decouples the indoor air volume from the high-mass granite walls, preventing the stone from acting as a natural heat sink. Consequently, any admitted solar or internal heat gains become trapped within the heavily insulated and airtight volume, leading to severe summer overheating.

(c)

Modern Case 2: The concrete-based reconstruction exhibits total thermal transparency with an immediate thermal response (0.0 h lag). This volatility is driven not only by the comparatively lower thermal mass of the concrete frame but primarily by the high window-to-wall ratio (WWR = 0.42). The large Southeast-facing glazing acts as a ‘Solar Trap’, admitting direct radiant heat that entirely bypasses the envelope’s limited inertia. This indicates that, in modern typologies, aperture geometry (window size and orientation) overrides material mass as the dominant thermal regulator, resulting in immediate temperature spikes in response to solar availability.

4.2.2. Internal Zoning and Hygrothermal Stabilisation

The monitoring data (

Table 5. Annual Hygrothermal Stability Metrics (Case 1: Pinheiro Novo).

Functional Zone	Mean Temp. [°C]	Stability Index [SI]	Min/Max [°C]	Annual Range [ΔT] [°C]	Mean RH [%] (Min/Max)
1st-Floor Bedroom (T1)	16.02	4.54	7.7/26.9	19.2	71.7 (44.8/91.5)
Ground Floor Storage (T3)	14.83	4.11	7.9/23.7	15.8	72.1 (48.0/86.1)
Outdoor ambient (T4)	13.90	6.55	0.6/30.3	29.7	68.9 (21.1/99.9)

) establish a performance gradient between the functional zones of Case 1 (Pinheiro Novo), validating the envelope scale of the bioclimatic framework. The Stability Index (SI), defined by the Standard Deviation (σ), quantifies the passive persistence across the building’s vertical section. This analysis strongly indicates how material composition, the assembly of high-mass granite and stone-on-earth floors, works in tandem with functional stratification to regulate internal conditions.

The data reveal that the building acts as a multistage thermal filter. While the first-floor living space reduces outdoor volatility by 30.7% (SI 4.54 vs. 6.55), the ground floor achieves a 37.3% reduction (SI 4.11 vs. 6.55). This performance advantage of the lower level over the upper level is the result of a synergetic “buffer effect” defined by four vital factors:

(a)

Compositional Ground Coupling: The ground floor’s stone-on-earth construction establishes a direct thermal link with the ground. This geothermal coupling provides a near-constant baseline temperature, allowing the floor to act as a thermal anchor for the entire structure.

(b)

Lithological Inertia: The 60–80 cm thick granite masonry provides high-density thermal mass. This composition functions as a “low-pass filter”, blocking high-frequency external temperature spikes and contributing to the 16.5 h thermal lag identified in the material-scale analysis.

(c)

Vertical Functional Stratification: The building operates through spatial hierarchy. The unheated ground floor serves as a volumetric buffer, protecting the upper living quarters from ground-level moisture and cold air infiltration. Simultaneously, the horizontal zoning places the bedroom between the kitchen and study, further isolating it from direct envelope interaction. Conversely, the upper floor acts as a “thermal cap”, shielding the storage area from direct atmospheric exposure and solar radiation through the roof. The efficacy of this stratification is evidenced by the suppressed Annual Range (ΔT) of the T3 storage zone (15.8 °C), which is nearly 47% lower than the outdoor environment (29.7 °C), demonstrating the protection provided by the upper-floor ‘thermal cap’.

(d)

Hygroscopic Stabilisation: The high-mass envelope and ground coupling also provide significant moisture regulation. While the outdoor relative humidity fluctuates across a violent 78.8% span (21.1% to 99.9%), the ground floor remains anchored around a mean of 72.1% with a significantly narrowed range. This shows the “breathability” and moisture-buffering capacity of the uninsulated stone assembly, which prevents the internal environment from reaching the saturation extremes seen outdoors.

Stabilisation is further evidenced by the Annual Temperature Range (ΔT). While the outdoor environment fluctuates across a 29.7 °C span, the ground floor is restricted to just 15.8 °C. This 13.9 °C suppression provides evidence that the traditional Stone-on-Earth typology is not merely a collection of materials, but a sophisticated, multiscale thermal regulator. By negotiating between material mass, ground contact, and spatial zoning, the envelope ensures a core of passive persistence that remains decoupled from the extreme environmental volatility of the mountain climate.

4.3. Seasonal Thermal Comfort Analysis

The impact of the settlement logics (Section 4.1) and envelope materiality (Section 4.2) is quantified in the seasonal comfort profiles shown in Figure 9 and Figure 10 and the statistical summary in

Table 6. Thermal comfort for selected bedrooms (3 case studies) based on the Portuguese Adaptive Comfort Model.

	Case 1: Pinheiro Novo		Case 2: Montesinho 2		Case 3: Montesinho
	Adaptive Comfort Limits Compliance (%)
	Compliant	Non-Compliant	Compliant	Non-Compliant	Compliant	Non-Compliant
Winter	0.00	100.00	13.93	86.07	7.81	92.19
Spring	0.00	100.00	21.88	78.12	72.97	27.03
Summer	59.49	40.51	94.30 1	5.70 1	52.97	47.03
Autumn	0.14	99.86	14.32	85.68	35.21	64.79

Note: ¹ 17 July 2024 17:00:00 to 10 August 2024 00:30:00 (Extrapolation T7 Bedroom air temperatures); ² 5 December 2024 16:30:00 to 23 January 2025 15:30:00. External temperatures extracted from Case 3.

. The monitoring data reveal a sharp performance inversion between the heating and cooling seasons, with distinct failure modes observed for each typology.

Figure 9 and Figure 10 illustrate the temporal evolution of operative temperatures against the adaptive comfort limits (Category II), identifying three critical performance trends:

(a)

Winter Disconnect: All typologies struggle to maintain habitability without active heating. Case 1 (Vernacular) records 0% compliance, thermally anchored to the ground temperature (~8–11 °C). Case 2 (Modern) achieves only marginal improvement (13.9%) despite insulation, while Case 3 (Renovated) remains low at 7.8%, indicating that envelope improvements alone are insufficient to overcome the winter thermal deficit.

(b)

Summer Inversion: In the cooling season, the performance hierarchy reverses. Case 2 achieves the highest compliance (94.3%), leveraging high permeability for night cooling. In contrast, Case 3 (Renovated) drops to 53.0%, exhibiting frequent overheating peaks as shown in Figure 10b where temperatures exceed the upper limits. Case 1 maintains a moderate 59.5%, primarily through passive damping.

(c)

Transitional Delay: A critical divergence appears in spring (May/June). While Case 2 quickly adapts to rising outdoor temperatures (21.9% compliance), Case 1 remains at 0%, denoting a significant thermal lag in the high-mass envelope.

5. Discussion: The Multiscale Logic of Vernacular Bioclimatic Strategies

This study investigated how the vernacular architecture of the Montesinho Natural Park (MNP) addresses environmental challenges through a layered, multiscale approach. By correlating geological availability, envelope inertia, and settlement morphology with long-term in situ monitoring data to define a system of passive persistence, the findings illustrate that this adaptation operates as a hierarchy of filtration rooted in the geological landscape.

Just as vernacular strategies in the Portuguese coastal regions adapt to a lithology-scarce environment through timber structures [51], the MNP architecture exploits the local lithology to provide an inertial resource (Geological Scale). This resource is utilised by the Settlement Scale to shield the microclimate (First Filter), enabling the Envelope & Material Scale to finally decouple the interior through thermal mass (Second Filter).

5.1. Geological and Envelope Logic: The Inertial Filter

While the geological scale determines the resource availability (granite), the thermal performance is actualised at the Envelope & Material Scale. The validation highlights that the resilience of these buildings is a synergetic result of material density and lithological inertia. The traditional granite masonry (Case 1) exhibits a 16.5 h thermal lag (τ), effectively shifting peak thermal loads to the nocturnal cooling phase. This prioritisation of thermal mass over thermal transmittance (U-value) aligns with findings in other Portuguese vernacular contexts, such as the Alentejo region, where high-inertia envelopes were found to be the primary regulator of indoor comfort [52].

This is in sharp contrast to the modern reinforced concrete reconstruction (Case 2), which is calculated with a lag of only 0.0 h. The comparison reveals a fundamental divergence in performance logic:

(a)

Case 1 (Traditional): Relies on “Inertial Delay”, prioritising long-term stability.

(b)

Case 2 (Modern): Exhibits aperture-driven volatility. The instantaneous thermal response (0.0 h lag) is driven by the synergy between the high WWR (0.42) and the corner exposure. With two external reinforced concrete walls and large glazing, the room suffers from multidirectional transmissivity. It gains heat rapidly through the glass during the day but loses it equally fast through the uninsulated corner and glazing at night, indicating that aperture geometry overrides material mass in this typology.

This underscores that replacing traditional thick-mass envelopes with uninsulated modern materials destroys the building’s inherent capacity to regulate thermal volatility.

To systematically correlate the identified architectural forms with the monitored physical behaviours,

Table 7. Synthesis of Multiscale Bioclimatic Strategies.

Scale	Vernacular Logic (Architectural Strategy)	Empirical Validation (Monitoring Results)	Bioclimatic Function (Mechanism)
Settlement	Topographical Siting: Locating settlements in “valley-folds” or mid-slopes rather than exposed peaks.	DTR Suppression: Pinheiro Novo maintained a DTR 20.5% lower (4.68 °C) than the exposed Montesinho cluster (5.89 °C).	Shielding: The settlement morphology acts as a “First Filter”, pre-conditioning the external microclimate.
Envelope & Material	Functional Stratification: The Stone-on-Earth typology; unheated ground floor coupled to the soil.	Stability Index (SI): The ground floor achieved an SI of 4.11, significantly damping external volatility.	Buffering: The earth coupling acts as a “Geothermal Anchor”, stabilising the base of the dwelling.
Envelope & Material (Inertia)	Lithological Inertia vs. Aperture Geometry: Case 1 relies on massive granite; Case 2 relies on glazing.	Thermal Lag (τ) Divergence: Case 1 (16.5 h) relies on mass; Case 2 (0.0 h) is driven by the Solar Trap (High WWR).	Filtering: The mass acts as a “Low-Pass Filter”, damping 47% of external volatility.
Envelope & Material (Zoning)	Functional Stratification: The “Sandwiched” layout; living core buffered by attic and cattle shed.	Winter Disconnect: While effective in summer, the internal partitioning leads to 0% comfort compliance in the bedroom during winter.	Spatial lag/Impedance: The zoning creates a buffer in summer but acts as a “Thermal Barrier” in winter, isolating the user from the heat source.
Building-to-Building Morphology	Compact Agglomeration & Shared Party Walls: Maximising structural adjacency within the cluster.	Microclimatic Stability: The 20.5% suppression of the external DTR (Section 4.1) quantitatively proves the cluster’s collective shielding effect.	S/V Ratio Reduction & Mutual Shading: Physically minimises the exposed envelope area to convective winter heat loss and provides aerodynamic buffering.

synthesises the multiscale bioclimatic framework. Crucially, this synthesis highlights the intermediate morphological scale, where individual buildings function as a cohesive thermal network. As established in Section 4.1, the building-to-building adjacency (shared party walls) and narrow street geometries provide collective aerodynamic shielding and mutual shading, bridging the gap between macro-settlement siting and micro-envelope inertia. This matrix elucidates that the “Vernacular Intelligence” of the MNP is not a singular feature but a cascading system of environmental regulation. By aligning the settlement-scale topographical shielding (DTR suppression) with the envelope-scale lithological inertia (16.5 h lag), the analysis reveals a coherent hierarchy of filtration that progressively stabilises the indoor environment against high-altitude climatic volatility.

The morphological divergence observed between the compact, high-altitude cluster of Montesinho and the linear valley layout of Rio de Onor highlights the critical filtering role of the Meso-scale. While their cultural origins may differ, the thermodynamic necessity of this adaptation is supported by modern bioclimatic research. As demonstrated by Fabbri and Tronchin [53] in similar European mountain-and-valley topographies, vernacular archetypes inevitably shift their compactness and form factors to satisfy localised energy balances dictated by microclimates. Consequently, while Montesinho relies on strict aerodynamic shielding through structural density to survive severe plateau winds, Rio de Onor utilises a valley-aligned street grid to actively negotiate complex solar angles and channel valley breezes. Furthermore, recent computational simulations of mountainous traditional streetscapes [49] support this approach, indicating that longitudinal street orientations are deliberately calibrated to maximise incident solar radiation along the corridor while concurrently maintaining critical wind channelling. Ultimately, this substantiates that traditional settlement morphology is not merely a cultural spatial arrangement, but a quantifiable, thermodynamically driven aerodynamic filter.

5.2. The Paradox of Stability and Habitability

The integration of the seasonal thermal comfort analysis (Section 4.3) reveals that passive persistence is not a universal solution, but a seasonally dependent asset impacted by specific morphological flaws.

5.2.1. Summer: The Inversion (Ventilation vs. Insulation)

The data reveal two opposing routes to summer comfort, identifying a “Permeability Divergence” between the typologies.

(a)

Case 1 (The Passive Base): The traditional dwelling achieves stability (59.5% compliance) through Inertial Resistance. Relying on the 16.5 h thermal lag of the granite envelope, it delays heat penetration. However, the monitoring reveals a limitation: the “sandwiched” internal zoning prevents effective cross-ventilation. Consequently, the building exhibits “Inertial Drag”: while the massive envelope suppresses peak temperatures during a heatwave, its sandwiched zoning prevents it from flushing accumulated heat at night, creating a slow but persistent warming effect during prolonged extreme thermal events.

(b)

Case 2 (The Ventilated Success): Surprisingly, the modern concrete building offers the most manageable summer conditions (94.3% compliance). This underscores the critical importance of Active Adaptive Behaviour. While the owners utilise the space on an intermittent basis (approximately fortnightly), their active management of the high WWR (0.42) during occupied periods becomes the primary asset. Occupants open the windows to induce cross-ventilation, flushing warm air instantly. This strongly indicates that, in this climate, the operational ability to ‘dump’ heat (high permeability) is often more valuable than the envelope’s ability to resist it (insulation).

(c)

Case 3 (The “Heat Trap” Risk): Crucially, the retrofitted building reveals the danger of combining internal insulation with restricted ventilation. The internal thermal plaster reduces the granite’s ability to act as a heat sink. This is severely exacerbated by the building’s intermittent occupancy profile; operational protocols dictate that doors remain sealed during vacant periods. Consequently, any admitted solar gains become permanently trapped inside the airtight volume. Combined with a tighter envelope and a severely restricted Window-to-Wall Ratio (WWR = 3.98%), the solar and internal gains are trapped inside. Empirically evidenced by the negative Absolute Damping (−3.10 °C), the room becomes a stagnant “thermos”, forcing indoor temperatures significantly above external extremes due to a complete absence of cross-ventilation.

5.2.2. Winter: The Three Modes of Failure

Conversely, the heating season exposes why distinct typologies fail to retain heat, even when occupied. The analysis identifies three distinct failure mechanisms:

(a)

Case 1 (The Zoning Failure): The 0% winter compliance indicates a compound failure. The issue is not just the uninsulated stone but the Morphological Disconnect:

• Lateral Disconnect: The 12 cm brick partition acts as a thermal break, decoupling the sleeper from the kitchen heat source.
• Vertical Loss: The uninsulated timber ceiling and slate roof facilitate a “Stack Effect”, allowing heat to escape vertically before it can warm the volume. Thus, the vernacular Thermal Crypt is created by the inability of the layout to distribute heat laterally and the inability of the roof to retain it vertically. Consequently, the sleeping quarters remain thermally anchored to the ground’s geothermal baseline (~8–11 °C).

(b)

Case 2 (The Solar Trap Vulnerability): Despite internal insulation and double glazing, Case 2 remains cold (13.9% compliance). The culprit is the high WWR (0.42) combined with an intermittent occupancy profile. While large windows admit solar gain during the day, they act as rapid heat-loss bridges at night. Because the owners occupy the dwelling intermittently and rely only on a mobile electric heater, the space lacks continuous thermal input. The heater lacks the radiative capacity to counter the overnight geometric heat loss, meaning the architecture works against the occupant.

(c)

Case 3 (The Geometric and Inertial Liability): The low compliance (7.8%) reveals a fundamental failure in the renovation strategy, where aesthetic spatial changes compromised thermal performance. This is driven by three converging factors:

• Loss of Buffer Zone: Unlike Case 1, which utilises an unconditioned attic space as a thermal buffer, the high sloped ceiling couples the conditioned volume directly to the external roof envelope. Even with insulation, the elimination of the “attic air gap” increases the rate of heat flux compared to the buffered vernacular section.
• Uncharged Thermal Mass: The unit lacks a continuous heat source (cooking/occupancy) to “charge” the granite inertia. Consequently, the massive walls act as a “heat sink”. During vacancy, the granite cools to the ambient baseline, and the internal insulation—insulating thermal plaster, a premixed mortar with expanded polystyrene (EPS) beads)—is insufficient to stop the walls from draining heat from the air.
• Volumetric Dilution: These losses are compounded by the expanded air volume (50.70 m³) of the extended height space, which dilutes the heating system’s output and encourages stratification. The result is a system-dependent typology. Without the passive protection of a buffer zone or charged mass, the building relies entirely on active energy. When that system is intermittent (vacancy), the indoor temperature collapses immediately.

5.2.3. Shoulder Seasons (Spring and Autumn): The Inertial Drag

The limits of the vernacular strategy are most visible during the transition seasons, specifically spring, where Case 1 exhibits distinct undercooling.

(a)

Case 1 (The Seasonal Flywheel): The massive envelope, saturated with winter cold, resists the outdoor warming trend well into June. As noted in the monitoring analysis, this creates a “Cold Lag”: the building remains cool during peaks but fails to warm up quickly enough during the transition. This is driven by a convergence of morphological and meteorological factors:

• Meteorological Obstruction: Climatological data [26] indicate that May is characterised by significant precipitation (63.4 mm) and cloud cover, reducing the direct solar radiation available to charge the thermal mass.
• Solar Access Deficit: Even when solar gain occurs, the unfavourable orientation (NE) and the internalised zoning prevent effective heat transfer. Consequently, the dwelling acts as a “heat sink”, thermally anchoring the occupant to the cold granite baseline rather than the warming ambient air. This illustrates that Vernacular Inertia is a “Non-Selective Filter”: without active management (apertures/vents), it indiscriminately dampens thermal change, delaying beneficial warming just as effectively as it blocks detrimental peaks.

(b)

Case 2 (The Diurnal Collector): Unlike the vernacular, the modern typology does not exhibit this specific inertia lag and undercooling phenomenon during the transition period (spring/May) despite being exposed to the same meteorological conditions. This divergence is driven by Responsivity:

• Solar Capture: Case 2’s high WWR (0.42) allows it to harvest diffuse solar radiation even during cloudy May days (“Greenhouse Effect”), breaking the thermal inertia lock.
• Volumetric Heat Capacity: The reduced wall thickness of the modern envelope requires significantly less thermal energy to elevate its internal temperature compared to the deep lithological mass of Case 1. Thus, while Case 1 acts as a “Seasonal Flywheel” (carrying winter cold into summer), Case 2 acts as a “Diurnal Collector”, resetting its thermal state much faster.

(c)

Case 3 (The Assisted Thermos Advantage): While the combination of internal insulation and restricted aperture (WWR 3.98%) creates severe overheating in summer, Table 6 reveals a striking anomaly: Case 3 achieves a remarkable 72.97% compliance during the spring transition. Unlike the deep winter, where intermittent heating fails to overcome the extreme cold of the uncharged mass (7.8% compliance), the milder spring baseline completely changes the operational dynamic. When guests are present, the centralised hydronic system requires significantly less energy to bridge the thermal gap. Once warmed, the internal insulation effectively traps this active heat generation, along with mild daytime solar gains. During spring, the “Thermos Effect” acts as a distinct advantage, maintaining the comfortable conditions established by the active system long into the vacant periods.

5.3. Implications for Retrofitting: Balancing Inertia and Connectivity

The divergence in performance between the traditional granite (Case 1) and modern concrete reconstructions (Case 2) highlights the risks of abandoning lithological inertia in high-altitude regions. To address the identified performance gaps, sustainable retrofitting in the MNP should follow three strategic principles:

(a)

Preserve the Lag: The success of Case 1 in summer strongly indicates that the 16.5 h thermal lag is a valuable asset. Retrofits must avoid interventions that decouple the internal space from this thermal mass, as seen in the reduced summer compliance of Case 3 (53.0%), where internal insulation created a “Heat Trap”.

(b)

Manage Aperture: The success of Case 2 in summer (through cross-ventilation) challenges the notion that thermal transparency is purely negative. Retrofits should integrate “Adaptive Apertures”—large openings for summer night-purging that can be thermally shuttered (insulated) during winter nights to prevent the heat loss observed in the modern typology.

(c)

Passive Connectivity and Volume Management: Energy efficiency in the MNP cannot be solved by envelope insulation alone. The “Zoning Failure” suggests that retrofits must establish Passive Thermal Connectivity. However, simply removing partitions risks expanding the heated volume beyond the capacity of the wood stove. Therefore, a sustainable strategy requires:

• Horizontal Insulation: Insulating the roof/ceiling plane to stop vertical heat loss.
• Indirect Convective Loops: Introducing high-level passive vents (transoms) above internal doors. This facilitates a natural thermosiphon, allowing buoyant clean warm air to circulate from the kitchen ceiling into the sleeping quarters, while keeping the combustion source structurally separate.

6. Conclusions

This study evaluated the bioclimatic efficacy of Montesinho Natural Park (MNP) vernacular architecture. By correlating settlement morphology, envelope stratification, and material properties with long-term monitoring data, the research affirms that the region’s environmental adaptation operates not through a single feature, but through the multiscale hierarchy of filtration proposed in the introduction.

Validation of the Multiscale Framework: Responding to the study’s primary objectives, the quantitative analysis is highly consistent with the “ecological intelligence” of the system across three distinct scales:

(a)

Geological Scale (The Resource): The statistical analysis revealed a direct dependency between settlement location and lithological availability. The analysis indicates that the vernacular construction is strictly “site-specific”, utilising the locally available high-density granite not just for economic reasons, but for its thermal persistence. This massive lithological resource provides the foundational inertia required to counteract the region’s climatic volatility.

(b)

Envelope & Material Scale (The Inertial Filter): Translating this material resource into a typological system, the “Stone-on-Earth” envelope functions as a low-pass filter. The monitoring data revealed that this specific assembly (Case 1), integrating massive walls with buffer zones, generates a 16.5 h thermal lag, a peak Absolute Damping of 3.80 °C, and a high stability index (SI = 4.11). This configuration successfully suppresses the annual temperature range by nearly 47%, ensuring internal stability during the critical summer months.

(c)

Settlement Scale (The Shielding Strategy): Critically, the study establishes that building performance is further amplified by the settlement scale. Topographical analysis revealed that the sheltered positioning of Pinheiro Novo acts as a primary buffer, suppressing the Diurnal Temperature Range (DTR) by 20.5% compared to exposed high-altitude clusters. This suggests that the vernacular system “pre-conditions” the microclimate before it interacts with the building fabric.

The Habitability Paradox: However, the seasonal analysis uncovers a fundamental limitation. While the vernacular system excels in summer (59.5% comfort compliance), the 0% winter compliance in Case 1 exposes a “Morphological Disconnect”. The “Thermal Crypt” effect is driven not only by the uninsulated mass but by a zoning failure: the internal layout and uninsulated roof structure decouple the occupant from the active heat source.

This contrasts with the modern reconstruction (Case 2), which displays aperture-driven volatility. Here, the Solar Trap of large glazing overrides the material mass, creating a system that is responsive to solar gain (beneficial for summer ventilation) but critically unstable during winter nights. Equally critical is the failure of the internally insulated retrofitting (Case 3), which exhibited total thermal transparency (0.0 h lag) and negative damping (−3.10 °C) in summer. Driven by restricted ventilation (WWR = 3.98%) and intermittent occupancy, the internal insulation decoupled the indoor volume from the stone heat sink, creating a severe “Thermos Effect” that trapped heat and exacerbated extreme temperatures.

Implications for Intervention: These findings challenge the direct applicability of standard modern solutions. The failure of Case 2 informs that replacing lithological inertia with lightweight materials creates dangerous volatility, while the failure of Case 3 warns against “blind” internal insulation that traps heat. This provides a critical warning for urban regeneration strategies: applying generic, high-tech efficiency measures without respecting the original ‘thermal mass’ and operational protocols of the fabric constitutes a form of ‘maladaptation’, reducing the building’s resilience to heatwaves. Consequently, future interventions must shift from a logic of simple replacement to “Balancing Inertia and Connectivity”. Strategies should focus on preserving the high-lag mechanisms that ensure summer survival, while establishing passive convective loops (via vents and roof insulation) to bridge the zoning gap that causes winter failure.

Limitations and Future Research Directions: One of the limitations of this research is that the empirical data are derived from a limited sample size of three monitored dwellings. This constraint reflects the severe logistical and socio-ethical challenges of conducting uninterrupted, year-long indoor environmental monitoring within privately owned, continuously occupied rural vernacular homes. Furthermore, rather than attempting to isolate a single geometric typology (such as the dominant Protruding Staircase form), these three cases were strategically selected to capture the evolutionary trajectory of the region’s building fabric. Case 1, preserved by the geographic isolation of Pinheiro Novo, serves as the pristine lithological baseline, representing the uninsulated thermal physics common across the historical stock. Conversely, Cases 2 and 3, located in the more accessible Montesinho village, represent the prevalent modern interventions—linear concrete reconstruction and the introduction of interior insulation—driven by contemporary socio-economic demands. Therefore, while the quantitative thermal values are specific to these dwellings, the comparative analysis between the pristine vernacular baseline and its relatively more modernised counterparts provides highly transferable insights into the bioclimatic vulnerabilities introduced by contemporary retrofitting in Mediterranean-Continental climates (Köppen Csb).

Furthermore, this multiscale methodology is not intended as a rigid model, but as an open framework for performance-oriented conservation and bio-climatic urbanism. While this study quantified the thermal efficacy of the physical form, the framework is designed for expansion. Future iterations should refine these findings by integrating landscape-scale ecology and socio-cultural patterns, ensuring that the quantitative “Performance Signature” remains connected to the qualitative human needs that sustain these resilient environments.

Final Synthesis: Ultimately, this research reflects that the “intelligence” of the MNP vernacular is not an abstract concept, but a quantifiable physical system. By elucidating that stability is generated through the precise coupling of the village (shielding), the wall (delaying), and the layout (buffering), the study provides a grounded framework for maintaining these distinct performance scales while adapting rural heritage to modern standards of habitability, ultimately serving as an evidence-based template for passive thermal resilience in an era of climatic uncertainty.