**1. Introduction**

Building categorization allows dividing the building stock into several homogeneous building groups according to certain key features such as construction period, building volume, material, etc. Archetypes or reference buildings could be selected from each building group to represent the most significant categories/typologies of the building stock. Obviously, this is only possible within certain assumptions and limits. Yet, these simplifications of the building stock are necessary for policy development and any other activity that aims at addressing the whole built heritage stock. By o ffering such archetypes, building categorization supports a bottom-up analysis of the building stock that allows an assessment of the energy consumption and potential conservation threats to a large building stock [1–3].

In the case of historic buildings, however, local influences on building typology due to factors like the evolution of the economic structure, population concentration, and di ffusion will challenge the generalizing approach of categorization [4]. The combination of all these factors results in an intricate history of design, construction, and renovation process of the buildings, which makes each historic building unique and hard to be grouped. However, climate change mitigation and adaptation activities require a certain generalization of the built stock.

The severity and impact of climate change were rigorously assessed in scientific literature. According to IPCC's (Intergovernmental Panel on Climate Change) Fifth Assessment report [5], the increase in global surface temperature by the end of the 21st century is expected to exceed 2.6–4.8 ◦C

compared to 1986–2005 in the most pessimistic scenario. Together with this temperature increase, extreme climate events are expected to occur more frequently [5,6]. In South Tyrol, climate change is clearly apparent in the increasing temperature and changed precipitation pattern. For instance, there would be more tropical nights (Nights during which the temperature remains above 20◦C) in summer and more precipitation in winter. Moreover, heat waves and extreme rain events would be more frequent [7]. It is also urgen<sup>t</sup> to decrease the greenhouse gas emission in the built heritage sector to mitigate climate change. One of the barriers to climate change mitigation is the incompatible retrofit solutions [8]. Climate change and incompatible solutions impose grea<sup>t</sup> challenges on the built heritage sector by increasing the risks of energy ine fficiency, indoor overheating, and moisture-related damage to the envelope [9]. To precisely identify the e ffect of climate change on the performance of retrofitted historic building, a three-year research project is being conducted, which includes four steps: (1) the identification of building categories and reference buildings (partly presented here), (2) the identification and assessment of present retrofit solutions, (3) the assessment of the combined impact of climate change and retrofit solutions, and (4) based on the results of the previous steps, suggestions for adaptation measures that are compatible with present and future weather. In this paper, the methodology to categorize historic buildings is presented. This will be the basis for further climate mitigation and adaptation studies.

Among the influencing factors, culture background, social customs, and most importantly the climate should be emphasized. Climate variability can impact culture, landscape, and human settlement [10–12]. Moreover, many studies confirmed the relationship between building characteristics and the local climate. In fact, the morphology, the position and size of windows, the wall material, etc. of historic buildings present climate-responsive features [13–16]. In Alpine regions, a wide range of landscapes and buildings evolved in the process of inhabitants' adaptation to local climate. They are a constitutive and essential part of the Alpine identity, sharing similarity in reflecting Alpine living. Building settlement form, construction technique, and other morphological or technical characteristics display the logic of climate adaptation [4,17]. For instance, the masonry is constructed with two external stone layer fillings with aggregates bonded with earth mortar and lime mortar to resist harsh external conditions; the compact volumes limit the thermal dispersion; the size and position of the windows are designed to minimize heat losses; the unoccupied attics reduce the heat loss through the roof thanks to the storage of hay or other fodder [18,19].

South Tyrol is a typical Alpine region in the north of Italy. It is characterized by its mountainous topography and diverse climatic conditions. Consequently, it o ffers a good scenario for the analysis of the relationship between climate and building typology evolution.

In summary, climate may have formed the typology of historic buildings in South Tyrol to some extent. Considering severe climate change in the future, historic buildings that were designed, constructed, and renovated according to climatic conditions in the past may be vulnerable to new threats, which will a ffect their conservation or performance in terms of indoor comfort and energy consumption. Conducting a categorization with a special focus on climate allows analyzing the historical interaction between climate and human dwelling activities and, accordingly, verifying the possible e ffects of future climate on historic buildings. Furthermore, archetypes representing the main categories could facilitate assessing the performance of the built heritage stock and planning the adaptation strategies in changing climate context. This study focuses on listed historic buildings [20], since they have the priority to be conserved and retrofitted, and specifically on residential buildings, as they represent the largest portion of the listed stock in South Tyrol and most parts of Europe [21,22].

### **2. Building Categorization: A Critical Review of Existing Methodologies**

Building categories enable grouping di fferent buildings that have similar or comparable features with the scope of being representative. The number of descriptive features depends on the number of target buildings, available building inventory, etc. There are no standardized characteristics; requirements and characteristics are selected for the purpose of the categorization.

In recent studies, one of the most common categorization targets was to support the assessment of the energy consumption or emission of the building stock (Table 1), i.e., to establish a stock energy/emission model. In that case, archetypes are created representing each category before scaling their energy use according to individual impacts to model the energy use of the entire stock. [23]. In the literature, energy use-related factors such as geometrical and thermal–physical properties of the building, the heating and cooling system, the climate zone of the building, etc. are used in categorization [2,24,25]. However, selecting all the variables that are significant for building energy performance is not feasible due to the data availability and the complexity of the energy model. Famuyibo et al. [2] attempted to define the key variables of buildings based on their impact on energy use (Table 1). Through multiple linear regression analysis, typical weekly occupancy pattern (heating season) (low/medium/high), internal temperature (◦C), immersion heater weekly frequency, and air change rate (ac/h) were selected from existing inventories because they are significant variables that influence the total energy use. However, it was found that, due to the limitations of the dataset (lack of data such as occupancy behavior), more than 60% of the energy use variation could not be explained by the model. Moreover, the first three of the significant variables were excluded since occupant-related variables were standardized in the operation of a reference building.

In the case of historic buildings, categorization aims to support not only energy performance assessment but also risk mitigation and the identification of retrofit solutions (Table 1). In some cases, it is used as a process to analyze historic buildings through identifying the vernacular characteristics, cataloging the materials in the di fferent construction periods, etc. [1,26]. Similar to non-historic building categorization, geometrical characteristics such as floor area and number of stories are adopted due to the general availability and their close relation to building energy performance. Thermal and hygrometric features such as construction materials are important for the preservation of heritage and the selection of retrofit solutions; therefore, they are generally used in categorization. In addition to that, the protection degree or other legislative requirements are included in some cases to present the historic significance or renovation limits of the buildings [27,28]. Construction period is selected because it reveals further information about building typology, construction materials, building equipment, etc. [3,25,27], thereby implying an analysis of the social, legislative, and technical impacts on building typology. Moreover, features on the settlement level could present the rooting of building stock. Montalbán Pozas and Neila González [1] suggested that categories of historic buildings should consider the sociocultural, economic, and historical contexts. They identified the building categories in a historic stock according to key features on four levels: territory, urban planning, architecture, and construction process, where features like the width and orientation of the streets, typical parceling of the blocks, etc. help interpreting the development and habitability problems of the stock.

There remains the problem of lacking data [23]. Since historic buildings have a complex history of construction and repairs, survey work covering the whole building stock is still infrequent. To avoid using deficient data, qualitative approaches are conducted in studies, such as expert evaluations, literature reviews, and on-site surveys [1,3,25,29]. For instance, due to the lack of adequate statistical data, the categorization of Hungarian stock was based on expert judgements [3]. A qualitative study could help understanding the building typology from a genealogy point of view, focusing on how the typologies evolved [30]. It helps linking the typology with its historic context.

Once the key features are selected, the category structure could be defined. There are two main category structures: flow structure and matrix structure, as shown in Figure 1. The category process of a flow structure successively divides the whole building stock according to selected features. The matrix structure is formed with two main key features. For instance, in the TABULA (Typology Approach for Building Stock Energy Assessment) project [31], building types (single-family houses, terraced houses, and blocks) and construction periods were selected as the two main features. Both structures have strengths and weaknesses. For the flow structure, it could include enough key features to establish detailed building categories, but the key features and intervals should be carefully determined since too many categories could be generated and some categories may not be representative. For the matrix structure, only two key features are involved in categorization; therefore, other features should be carefully added into the description of archetypes, without influencing the category results.


**Table 1.** Variables found in literature for building categorization.

**Figure 1.** Category structure: (**a**) flow structure; (**b**) matrix structure.

### **3. Proposed Methodology**

The methodology proposed in this paper was developed to prepare building categories for further risk assessment and adaptation planning, while permitting the possibility to analyze the relationship between climate and building categories which would provide knowledge support for further studies.

In order to identify the relationship between building categories and climate, the climate of South Tyrol was firstly analyzed and subdivided into homogeneous zones (Figure 2, step 1). In each climate zone, building samples were randomly extracted from the building stock (Figure 2, step 2a). Probability sampling was adopted in this study to ensure the representativeness of the sample despite the limited research resources.

At the same time, key features were defined according to the aim of the categorization (Figure 2, step 2b) through a literature review including categorization studies and studies on South Tyrolean residential buildings. Experts were consulted on whether the key features were representative and feasible to be used in this study. The criteria to select the expert panel were as follows: people who share an interest in the research project, and who have the knowledge of South Tyrolean historic buildings or have the experience of building categorization. In this study, the expert panel included three researchers based in South Tyrol with an expertise on energy renovation of historic buildings, as well as a local architect specialized in the conservation and adaptation of South Tyrolean heritage. Then, the defined representative features were collected for the building samples (Figure 2, step 3a), from available building inventories and the literature (step 3b).

After the dataset of key features was established, it was used in a flow structure to categorize the building samples (Figure 2, step 4). Eventually, the key features of the categories were statistically analyzed and compared among different climate zones (step 5a, Figure 2).

The results were interpreted with a qualitative study of South Tyrolean historic buildings, climate conditions, historic and social–economic events which influenced building customs, etc. (step 5b). By tracing the development of historic buildings, the relationship between climate and building categories was analyzed (step 6, Figure 2).

**Figure 2.** The methodology for categorization.

### **4. Categorization of the Historic Building Stock in South Tyrol**

### *4.1. Climate Zone of South Tyrol*

The whole region of South Tyrol covers 7400 km<sup>2</sup> with altitudes ranging from 190 m to more than 3000 m (Figure 3). The surface area below 1000 m above sea level (a.s.l.) is 14.1% of the total area, while the surface area over 1500 m a.s.l. represents 64.4% of the total area [36]. Due to the mountainous topography, diverse climate conditions exist. To analyze and subdivide the climate, climate data of different locations in South Tyrol are required. In this paper, climate data used were from (1) Provincia Autonoma di Bolzano Alto Adige (including data of 30 representative weather stations), and (2) results of the 3PClim project [37].

**Figure 3.** Elevation of South Tyrol (extracted from digital terrain model, http://geokatalog.buergernetz. bz.it/geokatalog/#!).

In this study, sub-climate types were defined according to criteria introduced below, which describe the similarities and distinctions in climate patterns. The climate zones were generated based on the results of the 3PClim project, where geostatistical interpolation methods were applied with the aid of programming software and geographical information system software [37].

The descriptive criteria were defined with the consideration of Köppen climate classification [38], which is a widely used climate classification system. The main parameters used in Köppen climate

classification are annual and monthly sums of precipitation, and annual and monthly mean temperature. The fundamental scheme of climate classification includes five major climate types (tropical, dry, temperate, continental and polar) covering the whole global climate. According to Köppen climate classification, the weather stations found in South Tyrol would fall into four different climate zones: Cfa, Cfb, Dfb, and Dfc. The differences between the four climate zones are shown in Table 2, and they are all temperature factors. However, the precipitation varies largely in South Tyrol from a regional point of view (Figure 4). Since precipitation has a significant impact on a building's hygrothermal performance, it is necessary to include precipitation in the climate zone definition in this study.


**Table 2.** Climate differences among four climate zones defined by Köppen climate classification. T—temperature.

**Figure 4.** Mean annual total precipitation of South Tyrol (reference period: 1981–2010, http://www. 3pclim.eu/).

In this study, the average temperature of the coldest month was used to divide the relatively warm zones from the relatively cold zones. To emphasize the impact of precipitation on climate classification, the median amount of precipitation of 30 representative weather stations was introduced as a criterion to differentiate between relatively dry and relatively wet zones. According to these two criteria, four sub-climate zones were defined (Table 3).

As shown in Figure 5, Zone I lays at the southern part of South Tyrol, covering regions with an altitude below 800 m. Zone I covers Val d'Adige that stretches from Salorno northward to Merano, and runs westward along Val Venosta to Naturno. In the east, it covers a narrow strip of low land along Valle Isarco. Zone I also includes the southern part of Val Sarentino that has relatively low altitude. The climate of Zone I is characterized by relatively warm temperatures and less precipitation. Compared to Zone I, Zones II and III have lower temperatures generally. Zone II distributes mainly in two parts: (a) the western part of South Tyrol, which includes Val Venosta and its side valleys such as Val Senales, Val di Trafoi, Val Martello, and Val d'Ultimo below 1300 m in elevation, and (b) the eastern part comprising the districts of Val d'Adige and Valle Isarco, where the altitude is around

600 m–1300 m, as well as Val Pusteria and its side valleys. The climate of Zones II and III differs in precipitation (Zone II has less precipitation). Zone III includes the vast highland in central and eastern South Tyrol. A fourth climate zone exists but is not included in this study due to its limited presence in the region.

**Table 3.** Climate differences among climate zones in this study.


**Figure 5.** (**a**) Climate zones in South Tyrol; (**b**) main valleys in South Tyrol.

### *4.2. Historic Residential Buildings in South Tyrol*

According to the 2017 population census of South Tyrol [39], the residential stock is composed of 225,483 buildings in total [39], 34,160 of which were built before 1919 and 14,840 of which were built during 1919–1945. These two parts comprise 22% of the total stock (Figure 6). Only 10.4% of this stock was retrofitted in the past 10 years (Figure 6). While energy retrofit represents an opportunity to reduce a building's operational energy and CO2 emissions, forecasted climate change might impose grea<sup>t</sup> risks on the hygrothermal performance of building constructions after retrofitting. For this reason, there is an urgen<sup>t</sup> need to analyze the relationship between building categories and climate, and to prepare archetypes for further performance assessment.

**Figure 6.** (**a**) Residential buildings by construction period; (**b**) residential buildings retrofitted during last 10 years by construction period [39].

Among the large residential stock, 4537 residential buildings in three categories (rural buildings, urban buildings, and nobility buildings, Figure 7) are listed as historic buildings under protection. Since rural farmhouses form the outstanding landscape of the Alpine space, they were selected to be studied under the category of rural buildings (Figure 7). In urban buildings, the trade–residential nucleus, the Portici house (Figure 7), was studied because it is the most important urban residence in the culture, social, and economic centers of the cities in South Tyrol. It appears in Merano, Bolzano, Egna, Bressanone, Vipiteno, and Glorenza. For rural farmhouses and Portici houses in each climate zone, building samples were randomly extracted ensuring a confidence interval lower than 15% and a confidence level of 95%, as shown in Table 4.

Even though the application of the proposed methodology is on listed historic buildings, it could be used for any historic building stock.

**Figure 7.** (**a**) Listed residential buildings in South Tyrol, and the building stock for categorization (data source: http://www.provinz.bz.it/kunst-kultur/denkmalpflege/monumentbrowser-suche.asp); (**b**) scheme of a Portici house (© Antonio Monteverdi, http://www.antoniomonteverdi.com/sito/?page\_id=1228).


**Table 4.** Information about building samples.

### *4.3. Key Feature Selection*

Our literature review showed that the key features used for any categorization should be selected according to the targets of the categorization. Therefore, features that are performance-related and potentially climate-responsive were selected to construct the flow structure for categorization. To better reflect the influence of climate, geographic condition, and historical context, key features were selected in three scale levels: settlement scale, building scale, and element scale.

On the settlement scale, key features included the compactness of the settlement and the number of the adjacent walls of buildings. Here, "compactness" describes the concentration level of the buildings; a "compact" type means that most of the buildings in this settlement are surrounded by close obstacles, while a "sparse" type means that most of the buildings are exposed to wind and rain without close obstacles. Close obstacles are defined as obstacles with a maximum distance of 25 m, which refers to the obstruction factor of 0.4 in EN 15927-3 [40]. The number of adjacent walls expresses the density of the settlement layout. The compactness of the settlement influences a building's resilience to extreme climates, while the number of the adjacent walls influences the energy use of the building and the indoor thermal comfort.

On the building scale, the typical Alpine building forms were considered. Geometrical and thermophysical-related features including roof projection area, floor number, window-to-wall ratio, and construction material were collected. Data were taken from existing GIS (Geographic information system) maps from GeoKatalog of Province Bolzano [41], external inspections, and photo evaluations. Geometrical features may result in different energy performance and thermal comfort according to the literature review, and construction materials present different behaviors in terms of moisture dynamics. The building layout, which indicates the distribution of functional space, was studied from the literature as supplementary information. The layout of residence space, farm space, and commercial space influences the heating setpoint, i.e., the heating schedule of the building spaces; therefore, it affects energy consumption.

Valuable building elements, which have historic, cultural, natural, morphological, and aesthetic value, were summarized from the literature because they are the essence of historic buildings and a crucial factor in retrofit decision-making. Any retrofit solution should be compatible with the heritage elements. Therefore, they influence the performance of retrofitted buildings indirectly.

### *4.4. Building Categories*

To define a reasonable number of building categories, settlement compactness, construction material, and the number of floors were used to construct the flow structure of the categorization, while other features were used as supplementary information. Therefore, 12 building categories, representing 81.6% of the building samples, are defined for further study (Table 5). All the key features are compared among different climate zones in the subsequent sections.


*Climate* **2019**,

*7*, 139

decoration (e.g., fresco painting, stucco), arcades, bay windows, wrought-iron rails, stone stairs, etc.

### **5. Results and Discussion: The Impact of Climate on the Development of Dwelling in the Alps**

In this section, the differences in the key features of historic buildings in three climate zones are presented, as a result of the quantitative study. To interpret and discuss these differences, we made use of the qualitative information resulting from the study of building history. Discussions focus on the differences that were historically influenced by climate to explore the possible role of climate in shaping the building categories.
