Urban temporary housing environments—from a systematic comparison towards an integrated typology

Housing plays a central role in everyday life and the fulfillment of human needs. Temporary housing demand can occur due to migration, extreme environmental events or personal decisions, and is expected to increase in the upcoming years. This study aims to create a general understanding of temporary housing. We conducted an integrated comparison of 66 international temporary housing examples via tabulation (table work), in an interdisciplinary manner considering details regarding built structure, open spaces, area, infrastructure, organizational and socio-economic aspects. It is the first time that a systematic comparison via tabulation (based on the approach of Braun-Blanquet) is used to classify temporary housing environments. The process is described in detail. The application of the systematic comparison creates a detailed typology that allows extensions and further differentiations. The types that emerge from the tabulation have specific concepts regarding the structural-spatial organization, technical infrastructure, and organizational matters, among others. The typology was further examined in the context of previous groupings in published literature. This novel approach of analyzing and structuring temporary housing offers a comprehensive perspective that can work as a universal understanding and language for precise communication among different disciplines regarding temporary housing.


Introduction
Temporary housing includes all forms of accommodation of individuals with non-permanent intentions (Johnson 2002), covering structures built to be temporarily on-site, as well as permanent buildings that are used for temporary stays. A rise in the demand for temporary housing is expected (Perrucci et al. 2016) due to various drivers, including the increasing severity of impacts of 'natural' disasters, the effects of climate change, and political conflicts (Black et al. 2013;Foresight 2011;. For some individuals, temporary housing represents a makeshift solution until permanent accommodation is reestablished or found (possible drivers may be external such as the above mentioned or personal decisions forced by biographical disruptions or uncertain future perspectives due to e.g. separation or illness). Others may freely choose this form of living (experimenting with alternative forms of living, relocation for higher education, etc.). The contrasting pair of temporary and permanent only expresses intentions. Experience has shown that neither the first nor the latter always apply as planned. For this study's purpose, we define a maximum (intended) stay/duration of 10 years as "temporary". Purely touristic or hospital stays are excluded since the study's object is housing which refers to settling and having everyday lives in the understanding of Katz (2017) and scholars of the subsistence theory (Mies and Bennholdt-Thomsen 1999). We chose the term 'housing' over 'accommodation' as the active verb to house stresses that it is constituted by various actions to fulfill human needs (Kennedy et al. 2008;Turner 1972). The study focuses on the urban environment, all analyzed examples are situated in or are suited for an urban context. 1 The research interest in temporary housing is rising. As highlighted by Albadra et al. (2018), the number of scientific publications on temporary housing has increased continuously over the last thirty years (see Johnson 2002;Félix et al. 2015;Perrucci et al. 2016;Al Asali et al. 2019). NGOs and public administrations have also engaged with the topic, releasing reports and guidelines suggesting procedures in post-disaster and migration events (Corsellis and Vitale 2008;IFRC and OCHA 2015; Sphere Project 2018; UNHCR s.a; UNHCR Shelter and Settlement Section 2016). Several authors from different backgrounds (post-disaster, migration, architecture) have tackled the challenge of classifying the different forms of temporary housing (Quarantelli 1982(Quarantelli , 1995Corsellis and Vitale 2008;Johnson 2010;Kikano et al. 2015;Albadra et al. 2018;Kronenburg 1995). Within and between different backgrounds ambiguous definitions and use of terms regarding temporary housing have been observed (Quarantelli 1995;Wagemann 2017).
This study is connected to a research project 2 that assesses and creates innovative and sustainable urban temporary housing concepts in an interdisciplinary way. 3 Hence, the study aims to create a clear understanding of the complex phenomenon of temporary housing, focusing on the integral parts that it comprises. Various scholars have addressed the challenges of interdisciplinary research (e.g. Borrego and Newswander 2008;Carr et al. 2018); for the aim of this study, lots of communication was necessary to create a common understanding of terms and concepts (Strasser et al. 2014).
To generate knowledge about temporary housing and its constituting parts, we applied a systematic comparison via tabulation (table work). With this method, we compared recent international temporary housing examples with each other to learn about their characteristics, similarities, and differences and eventually their principles and concepts. This method has never been applied to the field. We hypothesize that a systematic comparison via tabulation is a suitable method to compare temporary housing environments with each other. It is an integrated approach that allows the manifold disciplines to map the different elements of the housing environments and deduct a typology of temporary housing environments. The result is a typology that arranges the examples into comprehensible groups. Their designation provides explicit terms for the different forms of temporary housing, enabling clear communication within and beyond diverse research fields. Another benefit is that the typology is extendible providing an excellent basis for further research.
To understand the approach, we describe the process step-by-step in the method Sect. (2.2), after giving insights on the study's theoretical background (2.1). The results section is arranged according to the method's procedure: owed to the paper's brevity three out of the 66 examples and their associated groups are described (3.1 and 3.2). We then present the typology that emerged from the tabulation (3.4). This typology is then contrasted with four approaches of published literature (3.4) (2.2). In the discussion, the suitability of the method and results are discussed (4.1). Lastly, we review the method's challenges and prospects, concluding with an outlook on future research (4.2).

Theoretical approach
Out of manifold possible approaches we chose Gestalt theory to be the guiding metatheory of this research because it is integrative, interdisciplinary, and evidence-based (Schneider and Scharmann 2020). Gestalt theory proposes an integrated view on the topic by dealing with its constituting parts, stating that the whole is other than the sum of its parts (Neuweg 2019;Arvidson 2014), therefore, the whole (temporary housing) can only be understood by studying its inherent parts (building, area, users, etc.). This approach is suitable for complex subjects such as housing. According to Gestalt theory, systems cannot be forced into any arbitrary form. The measures of order have their origin in the system itself (Kriz 2008). The interpretation of the table resulting from the systematic comparison is guided by the characteristics of its structure.
A normative value within the study is to enable a good life for the residents, providing them opportunities to satisfy their needs (Max-Neef, 1991) and organize their everyday life in a self-determined manner (Mies and Bennholdt-Thomsen 1999). This premise is kept in mind throughout the whole study. This is part of the planning principle "Innenhaus-Außenhaus " (Hülbusch 1981), which addresses the enclosed (Innenhaus) and open spaces (Außenhaus) in context with people's freedom of action (Mayrhofer 2019). The assessment of private and communal enclosed and open spaces has therefore received special attention.

Method
Systematic comparison is a descriptive, comparative and synthesizing method that was originally developed in phytosociology to describe and compare plant communities with each other (Braun-Blanquet 1928;Tüxen 1970;Moore 1972;Dierschke 1994;Ewald 2003). Compositional patterns and gradients of research subjects are detected and analyzed in tables. Scientists from other areas applied the method to assess various research subjects, such as landscapes (Claisse and Géhu 1978;Theiling 1997;Pichler 2006). In landscape planning, it has been applied successfully since the 1990s for the differentiation of built and open spaces (Damyanovic 2007). In its original application, the (plant) species that occur at the assessed spot are recorded in a table. For applications in other areas, the characteristics depend on the research question posed (Theiling 1997): in landscape planning, all structural, material and socio-economic characteristics relevant to the research objective are recorded. The same applies to this case. The research project posed the question of innovative and sustainable temporary housing forms, which is why these characteristics got assessed (and are included in the table). We have not only selected innovative and sustainable examples, but the intention was to examine implemented examples for their sustainability and innovation potential. Additionally, relevant details on the built structure, area, user groups, duration on site, etc. were recorded. Figure 1 shows the process of the applied systematic comparison. After determining the aim, we collected a wide range of distinct contemporary temporary housing examples in the urban context (see (a) in Fig. 1). We recorded all examples for which enough information was available. Scientific journals and books, reports from humanitarian agencies and administrations, recommendations, and personal recollection (b) served as sources. A database was set up to systematically gather the examples' information regarding the structural and spatial environment, technical infrastructure, socio-economic organization, life-cycle considerations, and innovation potential. The examples helped to inform the compilation of attributes for the database (c). The minimum requirements for entering the database were the presence of a picture, the precise location, and basic information on the structural-spatial environment, technical infrastructure, and details on the users (d). For each example, a profile was created documenting broad information giving insights on its constituting parts (the use of materials, dwelling type, private and building-related open spaces, former land use, water supply, electricity connection, heating system, duration on site, user groups, as well as innovative aspects, among others) (e). Like other scientists comparing bigger numbers of housing environments (e.g. Seike et al. 2018), we had difficulties accessing data regarding all mentioned aspects. Only a few examples were well-described within a single source, therefore, YouTube videos, reports, news articles, documentaries, Facebook pages, Google Earth, communication with developers, and on-site visits provided additional insights. A total of 103 current examples from various countries (Europe, America, Asia, Africa) were collected from mid 2018 to mid 2019.
As preparatory work, we carried out a phenomenological comparison (Crease 1995f) (f) to perform a first grouping of temporary housing environments. For this, we took an image showing the entire dwelling of each example and compared them with each other, arranging similar examples according to their visual impression/appearance. Characteristics that were not shown on the picture (e.g. foundation, distribution of rooms, etc.) could not be considered in this approach. Like this, an initial sequence was created from individual units towards multi-unit buildings and from less durable (tents) towards more durable structures (permanent housing).
While comprehensive descriptions of the examples are not necessary for the phenomenological comparison, they are a prerequisite for the systematic comparison. This led to the exclusion of several examples in the database due to missing accessible information regarding specific components of the housing environments (g). Out of the 103 examples in the database, we assessed 66 via tabulation. In tabulation, the examples are described by the appearing characteristics, which we refer to as criteria. They are extracted from the assessed examples that were noticed during the survey were included in the database and subsequently the table. This corresponds with Gestalt theory's requirement: Arnheim expounded on the importance of taking everything perceived literally (Arnheim 1962). In our case, that means that all aspects perceived during the investigations entered the database and subsequently the table. The tabular comparison arrangement started by putting the first example in a column and the relevant criteria in the rows below, marking the intersecting field with an "x". The criteria are for instance "building has a durable foundation", "unit is lockable" or "user group students". The criteria are formulated positively so that only appearing characteristics (no lacks like "has no PV panel") are described. That leads to a first column full of "x". Incorporating further examples, we assessed all existing criteria, ticking if occurring or left blank if the criteria are non-occurring in a given example (i). If examples contained characteristics that were not mentioned in the table yet, we added the new criteria and re-examined the already incorporated examples for these new criteria. 4 The compilation of all analyzed examples amounted to 369 criteria that ensured that they were described in a multi-faceted and interdisciplinary way. The table is of a rather unwieldy dimension. Figure 2 shows a schematic diagram of the tabulation. The complete list of criteria can be extracted from Table S2 in the appendix.
Once the table was set up (raw table), several arrangement steps were applied: in the first step we organized the examples according to the order of the phenomenological comparison (h). This order is merely a presorting that works as guidance for this step and is moot for the further procedure. Then, we sorted the criteria by the number of fulfilled criteria placing the most frequently occurring criteria at the top of the table and subsequently all others in descending order (constancy table) (i) (Ewald 2003). We compared the columns and rows with each other and sorted them according to their similarities and differences (j). This was an iterative process, comparing the presence and absence of "x" until an order was established in which examples and criteria of similar patterns are placed next to each other.
The established order created a pattern (by accumulations of 'x' and unmarked spaces). A proper presentation of the sorted table is difficult due to its size. Figure 3 visualizes the emerged structure. A legible edition can be found in the appendix (Table S2). The resulting structure could then be detected (k): we identified and designated groups of similar structures. The groups hold criteria that are distinctive for one or various groups (character and differential species in phytosociology (Weber et al. 2000)). Each group is represented in the table by one or various examples that have similarities with each other and differ from the others regarding the occurrence and absence of criteria. The groups, so-called syntaxa, are organized on various levels in a typology following the systematic of phytosociology (m). From bottom to top, the levels are called types, alliances, orders, classes, and formations, each of them consisting of one or various of the previous (Schaefer 2012) (Table 1 in  3.3). Figure 4 shows a hybrid diagram that emphasizes the connection between tabulation and typology: the first five lines are the typology's syntaxa (more details can be found in 3.4.). Additionally, at the bottom, the names of the alliances are displayed. The table shows the selected criteria of the examples that prove to be the distinctive criteria of all syntaxa.
Literature offers defined terms and concepts such as the definition of a single-unit house in contrast to multi-unit buildings (Coolen and Meesters 2012;Harenburg and Wannags 1991). When creating the typology, existing empirical knowledge and defined terms could be taken as hypotheses and be examined (Fig. 1). The syntaxas' names represent the occurring phenomena of the respective group. Just as in phytosociology, the names are concise and carry an idea of the character of the syntaxa (Dengler 2016). The names cannot contain all characteristic aspects and be 'wholly adequate' (Weber et al. 2000, p. 740); however they must be explicitly making clear which form is meant. The syntaxa created in this study is described and explained in the results section. The arrangement of the examples and criteria according to the appearances and absences of "x" matches the organizational structure of Gestalt theory's principles of grouping (Wertheimer 1923) (step k). The structure created by the presence, co-occurrence, and absence of criteria had to be comprehended (n) and raised to a level of abstraction where concepts could be distinguished (o). The concepts then have to be understood in the context of the different forms of temporary housing.
The method of tabulation allows further table arrangements to be made to assess specific aspects. Section tables (Teiltabellen), for example, display selected syntaxa with their encompassing criteria arranged to thematic groups (Holzner et al. 1980). They show a clear depiction of the different disciplines ((p), in Fig. 1) and are a helpful tool to understand the diverse concepts that are inherent to the syntaxa. Table S1 is a section table that shows selected criteria of detached individual tents and detached durable multi-unit buildings. The criteria do neither hold the order that resulted from the tabulation process nor all criteria. Relevant criteria are arranged according to disciplines to show the properties of detached tents and detached durable multi-unit buildings regarding the built environment,  (Table S1) 1 3

Fig. 4
Hybrid diagram showing the connection between tabulation and typology spatial organization and environment, infrastructure, and circularity, socio-economic and organizational aspects. The apparent concepts of the two syntaxa are illustrated and discussed in the following chapter.
In addition to the presented method, literature screening was carried out. In the first step, we screened scientific databases (Scopus, Science Direct, Google Scholar). Whenever publications indicated further relevant publications, they were also assessed (snowballing). Several publications contain groupings of temporary housing environments, of which some are further analyzed and contrasted with our typology (see 3.4 and 4.1).

Results
In this chapter, the results of this study are presented. Since not all 66 analyzed examples can be presented, we describe three in detail in Sect. 3.1 to offer an impression of the examples and show how the descriptions got translated into criteria of the table. In 3.2 we describe the general principles of the example's types. In 3.3 we present and describe the whole typology and in the last Sect.3.4, we place it in context with groupings of published literature.  (Table S1 and S2) are italicized.

Example staff accommodation in Balukhali II
This example describes the staff's accommodation adjacent to a field hospital in Balukhali, Bangladesh in eary 2018, approximately six months after massive movements of people fleeing the Rohingya genocide started. The large number of people settling the area of Cox's Bazar necessitated provision of additional health posts and humanitarian assistance, thus the field hospital was established. The base camp for expat health workers in this field hospital consists of individual unit tents (criteria in the section table (Table S1): frame construction with lightweight walls, supported by poles, accommodation without foundation) (Figs. 5 and 6). The settlement is equipped with a field kitchen, on-site sanitation, and offgrid water supply (field kitchen in another dwelling, on-site sanitation (pit latrine), water supply by water trucking and drinking water supply via bottles). Energy provision is provided by generators (emergency power generator). The camp is organized on different layers: the kitchen, sanitation, and laundry facilities are to be used by the residents together, energy provision, and waste collection have facilities on the neighborhood level (groups of tents), the appropriation of the open space surrounding the units happens individually.

Example UNHCR family tent in Zaatari refugee camp
This example describes a specific accommodation in Zaatari refugee camp, Jordan, in the first years of the camp. The camp opened to host people fleeing the ongoing violence in Syria (user group migrants, user group fled from violence). In the camp, there were detached units without foundation made out of polyester-cotton, a lightweight flexible material, supported by poles (Fig. 7). They were individual units on one plot without physical borders like fences, the units have private open space through appropriation (for 1 3 cultivation, drying laundry, play, etc.). The technical and social infrastructure was organized by the camp (Fig. 8) and developed rapidly, just like the settlement. The situation today is rather different. In 2013 there were around 8,000 tents with either shared or private sanitation facilities. Clean water was trucked to the camp and wastewater was then removed by trucks again. The camp was connected to the electricity grid, units could be supplied with energy as well.

Example Home 21 in Vienna
This example includes two durable 4 (in some parts 5)-storied buildings in Vienna, Austria. Primary materials are concrete and steel (temporary housing environment (THE) made out of heavyweight, rigid materials). The buildings include around 250 apartments for people entitled to social housing and people with an increased need for support (first housing for homeless persons, single-mothers, etc.) (user group: locals, user group: people entitled to social housing, user group: homeless persons, among others). The ground floor includes wheelchair-accessible flats and commercial units (mixed use in THE). The apartments are fully equipped and connected to centralized technical infrastructure (connection to power grid, unit connected to centralized water supply, THE has sewage connection). The residents have private balconies (unit contains private open space (balcony)) ( Fig. 9) and collectively used open space on the ground floor (Fig. 10). The buildings have been developed within a city's subsidy program to alleviate housing shortage, encouraging the construction of temporary low-cost housing in areas that are eventually intended for other uses (subsidized living units/state-funded, mixed building land (zoning plan)). They are intended to be used for housing for 10 years and afterward for commercial uses. Since the example, like others, was evaluated via desktop research, limitations and errors can occur. As it could not be discerned on maps and plans, the exact location of the building's entrances is unknown.

Step 2: …to principles and concepts of the syntaxa
The examples have specific characteristics that have to be abstracted and elevated to a level of principles to be compared with each other.  Table S1 shows that there are criteria that are typical for the alliances, for detached individual tents for example such as 'unit accessed by footpath' and that there are clear differences between the two types T1 and T2, such as 'unit embedded in publicly accessible urban fabric'. But there are also gradients within the syntaxa: criteria that occur in one and the other demonstrating the transitions from one to the other. The differences regarding

Detached individual tents (A1)
The analysis of the systematic comparison (Table S2) and section table (Table S1) Table S1). Cooking, sanitation, and laundry facilities may be organized on unit, neighborhood (group of units), or camp level. In camps, the provision of tents, facilities, infrastructure, and services is generally centralized. Tents are mostly placed on unbuilt green land that is not formally parcelled out, and the residents do not own the land. When comparing this alliance with others in the table, the absences are apparent: the units are not lockable, which indicates for a lack privacy and security. Since the residents are not owners of the plot, and not always of the tent, they have (limited) entitlements. Figure 11 shows the physical concept of detached individual tents in camps (T2).

Detached durable multi-unit buildings (A11)
The buildings of this alliance are placed on durable foundations, an indication for being residential buildings planned to be permanently on-site. They are made out of heavyweight, rigid materials or shipping containers and all have thermal insulation. The buildings have various apartments per floor (Fig. 12) and are multi-storied. The housing environments may consist of more than one building and often comprise over 100 units in total. The buildings are connected to centralized infrastructure (unit has sewage connection, unit connected to centralized water supply, connection to central power grid, municipal waste collection). The units work as independent households (unit has private toilet, unit has private bathroom, private kitchen) but share some facilities (shared laundry room). The units are lockable, an important condition for privacy and security. Since the residents are renting they have (limited) entitlements.

Typology
The systematic comparison via tabulation created a table in which the examples were compared with each other according to differences and similarities in 369 criteria (Fig. 3). This structure led to a typology of urban temporary housing environments on five levels creating 2 formations, four classes, fourteen orders, twenty-two alliances, and twenty-six types ( Table 1).
The first level of structuring is into two formations (Table 1): demountable housing environments of planned temporary presence (F1) and temporary housing environments in durable buildings (F2) that distinguish between the presence and kind of foundation. With the term 'foundation', we refer to buildings' elements that transmit structural and externally applied loads to the building (wind, vibrations, seismic shocks, and tremors), directly to the ground (Jumikis 1984). Only elements that are anchored to the ground are considered foundations. 5 The assessed examples have either no, a demountable (removable without major impact to the ground) or durable (made out of concrete, removal by heavy machinery, the ground is impacted for longer) foundation. The absence or presence and kind of a foundation is an indication of the temporary nature/permanence of a housing environment and has an impact on its subsequent use or re-use.
Demountable housing environments of planned temporary presence (F1) encompass dwellings that are built for a limited presence on a particular site. We call them demountable because these structures are modular and/or easily dismantled. The examples of this group are further arranged in 2 classes: accommodations without foundations intended for residential use (C1) and buildings on demountable foundations intended for residential use (C2).
Accommodations without foundations intended for residential use (C1) were further structured into: tents (O1)-textile dwellings supported by poles, cabins (O2)-hard shell units (like a in Fig. 13), and mobile accommodations 6 (O3)-dwellings that are moveable

Table 1
Typology of urban temporary housing environments (the allocation of the three examples discussed in 3.1. is highlighted in grey) *These types are formulated as hypotheses since they contain less than 4 examples without lifting, which means that they include their mode of transportation already, (on wheels or floating on water such as b in Fig. 13).
Buildings on demountable foundations intended for residential use (C2) were structured by the number and settings of their residential units. Individual unit buildings on demountable foundations (O4) have one unit per entrance, multi-unit buildings on demountable foundations (O5) (such as c in Fig. 13) contain various units per entrance, and halls on demountable foundations (O6) a large room with one (or several) entrance(s) containing a large number of sleeping niches or bunks instead of units. 7 The other formation, temporary housing environments in durable buildings (F2) include all buildings on durable foundations (coinciding often with heavyweight materials). 8 Accommodations of this formation are intended to stay on-site for a long time ('permanently'). They may have been built for temporary residential use or another purpose and are (temporarily or permanently) used for temporary accommodation.
Two classes are differentiated: Buildings on durable foundations intended for residential use (C3) include buildings that were constructed with the intent to accommodate people, whereas non-residential buildings on durable foundations (C4) include buildings that were constructed for other purposes. Buildings on durable foundations intended for residential use (C3) have in common that they are built for accommodation purposes, so their facilities enable habitation without major structural modifications. They are structured into multiunit buildings on durable foundations (O7), reused social facility buildings (O8) (like d in Fig. 13), and reused hotel buildings (O9). In the latter two cases, the buildings are utilized for other uses than what was initially intended (social facility and touristic stays).
Since the utilization of non-residential buildings on durable foundations (C4) as accommodation for people represents a repurposing, also referred to as adaptive reuse, it entails that they have to be (structurally) adapted or complemented. We assessed repurposed administrative buildings (O10), repurposed office buildings (O11), repurposed retail buildings (O12), repurposed industrial buildings (O13), and repurposed educational buildings (O14).
The orders were further organized into alliances according to the urban layout (detached, semi-detached and attached accommodations describe if the dwellings share walls with neighboring dwellings or not) and types according to the settlement properties (in a camp 9 or within the urban fabric of a city).

Typology in the context of published literature
In published literature, various groupings of temporary housing environments can be found. The approaches are determined by the particular research questions or applications. NGOs and public administrations, for instance, often consider different forms of assistance (such as the distribution of material, financial support, etc.) in addition to accommodation options (Sphere Project 2018; UNHCR s.a.), and scientists create typologies according to their specific field of study. Our research question was informed by an interdisciplinary project pursuing a general approach. Table 2 shows the structure and concepts of four grouping approaches (Quarantelli (1982), Kronenburg (2014), Johnson (2010), and UNHCR (2018)) and an overview of our typology. The four approaches were selected since they focus on temporary housing solutions, mainly excluding other types of aid (financial support, food supply, etc.). Quarantelli's (1982) approach structures housing environments that are used in postdisaster housing according to the planned duration and time of application. Kronenburg (2014) applied a grouping of "mobile and temporary building systems" according to their number of components, as well as the time and effort needed to assemble the buildings on-site. Similar to other authors (e.g. Onay 2018), Johnson (2010) differentiates post-disaster housing into two main groups: types that need new construction 10 and types that do not. Additionally, she diversifies according to organizational, built, structural-spatial, and social-economic aspects. Compared to Johnson's approach, in our typology the structuralspatial conditions of housing environments feature prominently in contrast to the socialeconomic aspects. UNHCR's (2018) typology of shelter types for refugees is solely based on structural characteristics of the housing environments. 11 It appears that after the first grouping into three divisions (similar to ours), known building types are applied to designate the diverse structures.
A common feature of the approaches by Johnson, UNHCR and ours is that they differ at various (two or five) levels, on each level additional information is provided and different concepts can be detected. Since instructions on their derivation have not been published, questions regarding the applied methods for comparison remain open. Furthermore, we screened publications of various authors that compared temporary housing examples (Seike et al. 2018;Toma 2018) or case studies (Kelman et al. 2011) with each other, but did not find descriptions of how classifications were derived. It is a unique feature of our typology that we present the typology including its derivation.

Typology as the basis for a better understanding
Wagemann (2017, p. 235) discusses the difficulty "to find definitions that apply in different contexts because each disaster, each country, and the way houses take shape is different". To understand the phenomenon of temporary housing, clear concepts and consistent terminology are needed (Quarantelli 1995;Wagemann 2017;Pezzica et al. 2018). These should be independent of the trigger and generally applicable.
In Table 2 we introduced four published approaches to structure temporary housing environments and compared them with our typology. While the four analyzed approaches are suitable for the objective of their respective application, their suitability for a general application is varying and depends on the structuring features. Quarantelli's approach (see at the bottom of Table2) is difficult to apply to housing environments that emerge for other 1 3 reasons than emergencies and have unpredictable life spans due to its ambiguity. Typologies structured by material-structural characteristics are easier to apply to other contexts, though some might need adaptations. To apply Kronenburg's approach in general contexts, it would first have to be expanded for buildings other than mobile. Although there are similarities in the essence of Johnson's, UNHCR's and our typology, the typology we developed involved an interdisciplinary analysis, providing detailed wording and descriptions which allow a precise classification that is not restricted to the dwellings but also touches upon their environments. Our types include sound information on structural and spatial conditions. From the type designation detached durable multi-unit building within the urban fabric, it can be discerned that the dwelling is intended to be on-site for a longer period, since it is a durable building. The number of units indicates the density. It is surrounded by open space and has no connection to other buildings (detached) which makes its application more independent from the surroundings. Its location in the urban fabric indicates its possibility of connection to the surrounding social and technical infrastructure, as well as being bound by the city's regulations.
We take tents as an example to compare the different approaches' designation/allocation. They can be found in Johnson's and UNHCR's approach with the respective additional information (grouping) that they are non-permanent structures (UNHCR) or rather newly constructed (Johnson). Kronenburg does not assign it directly, the definition of relocatable buildings (transported in parts, fast assembly on-site, etc.) or demountable building (transported in many parts (..), compact for transportation) may both be suitable. In Quarantelli's approach, they may be allocated to any of the first three groups, depending on the duration of use (in the publication they are mentioned as temporary housing example). In our typology, there is the possibility to refer to them as tents but also to name them more precisely (A1-2, T1-3). The allocation to the formation and class states that the dwellings are intended to be temporarily on-site and have no foundation. Compared to the other approaches, our typology is unambiguous and more informative as it goes into greater depth.
Our proposed typology for the context of temporary housing environments offers the following advantages: a general applicability due to its structure, a clear wording and terminology, and the possibility to replicate and expand (see 4.2), in case further differentiation is desired. We suggest taking the typology's sytaxa (designations) to refer to the different forms of temporary housing. Applying a well-defined terminology based on tangible characteristics (spatial-built) offers a precise language for communication within and among the diverse fields and backgrounds. Only based on a common understanding overarching the different fields, a discourse between the disciplines can be created, facilitating an exchange of experiences, findings, approaches, and innovations.

Tabulation and typology as a basis for further research
One of the study's purposes is to understand temporary housing by assessing a variety of different housing forms. The systematic comparison leads to a 'qualitative type-creation according to characteristics' (Hülbusch 1981), based on the precise assessment of the criteria and their interrelations. It is the first time that this method is applied to the field of temporary housing. In section tables, the syntaxa's characteristic criteria become evident (by appearance or absence of criteria). This allows clear insights into the syntaxas' differing concepts regarding the built environment, spatial organization and environment, technical infrastructure, and socio-economic and organizational aspects. The conditions of the Table 2 Selected approaches to structuring temporary housing environments from published literature compared to our typology ¹For reasons of space limitations a reduced version of the typology is shown here (without types) physical and organizational environment are preconditions for the resident's freedom of action (Mies and Bennholdt-Thomsen 1999). Due to the thematic focus of this paper, these aspects have not yet taken center stage but should be examined in further research.
Further compromises, challenges, and drawbacks of the study are addressed in the following: It is the theory's (Gestalt theory) and method's (systematic comparison) requirement to depict the subject as a whole. Assumptions were made in cases where no public data was available, like surmising that Viennese households are serviced by the municipal waste collection, although the example's description does not explicitly state the fact. Ideally, all information can be verified. Since on-site visits were not possible for all examples, we mostly depended on external sources that do not represent the object itself but are the perceptions and reproductions of others. Biases (such as judgemental perceptions) may be included, which must be taken into consideration.
The examples found in scientific publications, newspapers, and on websites may overor underrepresented certain forms of accommodation. Specific triggers may lead to special housing forms, such as staying at a relative's place (e.g. after natural disasters) or self-built accommodations (e.g. in cases of homelessness), for instance. These kinds are underrepresented in literature and publicly accessible sources (Quarantelli 1980). Therefore, we stress that the typology represents an overview of the analyzed housing environments but cannot be considered to be all-encompassing.
The limited number of examples that could be collected in the required quality presents a challenge for the processing and interpretation of the table. To formulate exact subdivisions, more examples would be needed. For this reason, some syntaxa of the typology are to be understood as hypotheses, which are to be examined based on further examples and investigations (groups containing less than four examples are marked with an * in Table 1). Since the theory and method of the approach allow additions and differentiations, the typology can be expanded at any point.
The analysis indicates that the housing environments' diverse aspects affect one another (e.g. aspects of built environment affect infrastructure connection and circularity aspects, just like the organizational aspects affect the built environment and other areas). Besides the resulting groups of the typology, other gradients in the analysis table can be observed: multi-unit buildings on demountable foundations and multi-unit buildings on durable foundations, for example, belonging to different formations have various criteria in common. The assessment of these aspects provides in-depth information on the forms of temporary housing that is yet to be further detected and discussed.
This study could be deepened in two directions. On one side, the examples could be investigated in more detail (through on-site visits and interviews), on the other side, more examples could be assessed. Both would improve the results and may alter them. To expand this research by adding examples, the Table S2 can be taken as a basis. It is recommended to gather all information in a textual description (including pictures, maps, and plans if available). The gathering of the example's description should be driven by its properties. Additionally, the

Concluding remarks
This study had two main aims: to create a clear understanding of temporary housing environments and to demonstrate the application of the systematic comparison to temporary housing.
Recent international examples of temporary housing were analyzed and compared with each other via tabulation. This approach is novel to the field. We showed that the systematic comparison meets Gestalt theory's requirements of a complete depiction of the whole subject by assessing different aspects of the research subject. The systematic comparison led to a typology that classifies the housing environments according to their qualities on different levels.
Contrasting our typology with others, we showed that structural-spatial characteristics are suitable to structure the variety of temporary accommodations. The precisely defined terms facilitate a clear understanding of planned and existing structures. The suggested terminology provides a common language to scholars and involved persons of different backgrounds, promoting communication, understanding and learning among different disciplines. This creates the possibility to exchange knowledge and findings, fostering new findings and synergies of various disciplines. The various disciplines might learn from the experiences and structures of other fields, opening spaces for innovations.
The extensive research that has been carried out is a suitable basis for further investigations. The typology is not a rigid structure: the systematic comparison via tabulation allows the integration of further examples and criteria. Continuing the process started here, extensions (new types) as well as more specific differentiation of the classification (sub-types/variations) can be achieved. The method also opens up the possibility for in-depth investigations on examples and/or types, about which there is still a lot to learn.

Supplementary information
The online version of this article (https ://doi.org/10.1007/s1090 1-020-09812 -x) contains supplementary information, which is available to authorized users.Acknowledgements We thank our entire project team for the inspiring discussions, with special thanks to Gaetano Bertino and Thomas Märzinger, who collaborated in the collection of examples and creation of the database. We sincerely thank the three anonymous reviewers whose suggestions helped to improve and clarify the manuscript.
Funding The research of this project has been funded by the Vienna Science and Technology Fund (WWTF) through project ESR17-010 and by the Institute of Landscape Planning, University of Natural Resources and Life Sciences Vienna, Austria. Open access funding provided by the University of Natural Resources and Life Sciences Vienna (BOKU).

Data availability
The datasets generated during and analyzed during the current study are available from the corresponding author on request. A list of references used to create the database is attached to the submission and may be published as an electronic supplementary material.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.