The Multi-Criteria Scenario Framework has been methodologically validated on the basis of a combination of methods. An in-depth case study analysis is used as input parameters, allowing for testing of data and respective results (section 2.1). The input parameters are methodically combined within the framework (section 2.2), and it is tested which ‘if-then’ interactions can be carried out on the basis of scenarios for decision makers (section 3). By supplementing or adapting the energy targets and criteria, this exemplary approach can also be implemented in other case areas. The framework is assessed both in terms of content with respect to the results on spatial renewable energy targets and spatial species protection (section 4.1), and in terms of its applicability (section 4.2) (see Figure 1).
2.1 In-depth case study analysis
2.1.1 Research design using case study analysis
As an input parameter for the Multi-Criteria-Scenario Framework, an in-depth case study has been carried out following Yin, Yin [60]. A case is characterized by a contemporary phenomenon in a real context, where the boundaries between ‘phenomenon’ and ‘context’ are often blurred [60]. Here, the interrelationships between spatial categories for renewable energy and species are related to ‘phenomena’. These are framed by the external conditions of achieving respective renewable spatial target setting. The context is therefore essential to understanding the case
[61]
. In this way, theoretical and practical conclusions can be drawn about possible land use trade-offs, and the applicability of a Multi-Criteria Scenario model for tackling these in decision-making.
For the selection of a case study area, specific criteria were established based on a literature search on barriers in wind energy development [7, 62–67, e.g., 68]:
- Wind energy planning should take place at a higher planning level than local approval level, such as regional or state level planning for zoning, to explore possible trade-offs between (energy, species) targets and land use allocations. Regional planning is accorded an integral role in the overall coordination of land use planning [67, 69].
- A trade-off situation between land allocation and renewable energy use would allow the application of a Multi-Criteria Framework to be explored in a solution-oriented manner [68, 64, 66, 65].
- (Geo-)data on the planning situation should be publicly available, if possible, to be able to perform own area analyses in the model framework [cf. 70].
The Havelland-Fläming region in Brandenburg, Germany, was selected as a case study on this basis. Havelland-Fläming is one of five regions in the federal state of Brandenburg, located southwest of Berlin, and organized as a regional planning authority [71] (Figure 2).
2.1.2 Case study characteristics
The study area features regional wind energy planning where decision makers face the challenge of implementing wind energy spatial targets while balancing other land use interests such as settlement buffers, landscape protection areas, and forestry [72, 73]. Through zoning, regional planning creates perspectives on land use. It resolves stakeholders, power structures, political/economic/social/environmental uncertainties, and competing values
[40]
. In Germany, regional planning operates at the interface between the land development objectives of the federal states and those of the municipalities [68]. In Havelland-Fläming, which is in the process of updating their regional plan, draft regional plans have been revoked twice by the courts due to insufficiently justified planning criteria for wind energy [74][2]. As in 2022, a third plan has been developed and is subject to public participation [77, 78]. The majority of the planning documents have been made available to the public as well as geospatial data [77].
The selection of areas for wind energy in this case, is characterized by an exclusionary planning approach (e.g. excluding landscape protection areas, nature reserves, settlement buffers, among others) [77]. As a result, 1.67% of the region’s remaining area has been identified as suitable for wind energy, which hereafter will be referred to as ‘wind energy areas’. This share of the remaining space does not reach the statewide spatial target for the development of onshore wind energy as was set by the federal government. According to federal law (Wind Energy Demand Act), the state of Brandenburg must designate an area of 2.2% within its regional plans by 2032[3],[4]. Since this missing delta (approx. 0.53%, i.e. 38 km2) for wind energy sites has yet to be identified, this situation provides a suitable basis for a scenario approach. It suggests that there is a need to balance and weigh trade-offs, which is particularly relevant when evaluating how a Multi-Criteria Scenario Framework can help decision makers to quantify goal-oriented impacts of planning decisions.
For ground-mounted PV, a distinctive planning concept is lacking as this step is left to municipalities in this state and region, a subsequent planning level. Yet, basic provisions for PV relate to using agricultural land for technologies, such as agri-PV (agricultural PV). Agri-PV allows for the parallel use of agricultural land for energy and food production [78].
Unlike other German states, federal offices in Brandenburg have not introduced bird priority zones as a planning category
[cf. 79, 46]
. Species protection issues have been dealt with through the use of species-specific buffer zones
[80]
. Buffers were established for bird species, such as Red Kite (Milvus milvus), White Stork (Ciconia ciconia), Osprey (Pandion haliaetus) and other migratory and resting birds. However, this approach requires high quality data of breeding sites as well as scientific evidence for buffer estimates. It also mainly addresses sedentary species, such as Eagles, that rarely change breeding spots over time [81, 82].
2.2 Multi-Criteria Scenario Framework
A Multi-Criteria Scenario Framework for an explorative ‘if-then’-analysis includes input parameters as shown in Figure 3, which are based on the case study area. The model runs different scenarios that consider different approaches to bird priority zones combined with planning criteria used within the case study for wind energy and PV allocation (‘if’), and shows how these scenarios affect the potential of achieving renewable energy targets (‘then’). Specific data is required for these input parameters of the model, which are evaluated in the following sections (Figure 3).
2.2.1 Variations of bird priority zones approaches
2.2.1.1 Assumptions and target species for bird priority zones
The aim of the species priority zones is to protect functional areas for wind energy sensitive species early on at the higher planning level, rather than only at the subsequent wind energy permitting level (i.e. population-based conservation approach) [43, 79]. As there is no single approach for priority zones yet, different approaches have been examined in this analysis, e.g. in relation to target species and spatial designation approaches [46, cf. 79, 43]. The aim was to enable a discussion on which approaches can have which land footprints on wind energy and PV development.
Four premises were used to select the target species based on a literature analysis [44, 50, 48, 83–87, 45]: 1) Species priority zones would only be meaningful for those species that are considered sensitive to wind turbines. 2) Species must be widespread, with well-defined habitat parameters and a good knowledge and data base should exist. 3) Species must also be difficult to manage, for which bird priority zones add value, and have a relatively high variability of breeding sites, for which only species-specific buffers are less suitable. 4) Conventional protection and avoidance measures usually do not provide satisfactory solutions (see Supplemental Materials).
Based on these premises, eight raptor species were identified as suitable for species priority zone approaches: White-Tailed Eagle (Haliaeetus albicilla), Osprey (Pandion haliaetus), Red Kite (Milvus milvus), Common Buzzard (Buteo buteo), Black Kite (Milvus migrans), Marsh Harrier (Circus aeruginosus), Honey Buzzard (Pernis apivorus), and Hobby (Falco peregrinus). The availability of point data, which represent nesting sites, in the Havelland-Fläming region ultimately limited the choice of species. This limitation resulted in bird priority approaches for only two of the eight species originally considered for this analysis, the Red Kite and the Osprey. Although Red Kite populations are widely distributed throughout Europe, about 50% of the European population reside in Germany. As a significant portion of the population occurs in Germany, the conservation of Red Kites is considered to be a ‘special responsibility’ [83, 88]. Ospreys are classified as vulnerable in the Red List of breeding birds, and are also considered as sensitive to wind turbines [89, 90].
The Multi-Criteria Scenario Framework therefore uses three differing bird priority zones for testing as input parameters: Two for the Red Kite (Milvus milvus), including the Top 5 and Top 10 most suitable habitats, and one for the Osprey (Pandion haliaetus), which depicts its Top 10 most suitable habitats. The bird priority zones were developed on the basis of habitat modeling by ARSU GmbH in collaboration with the research project ‘Bird Priority Zones for Species Protection’ at the Berlin Institute of Technology (TU Berlin), which was sponsored by the German Federal Environmental Foundation (DBU) [91]. In addition to these three priority areas, an aggregation of all bird priority zones was included in this analysis, as well as only those bird priority zones that overlap with each other, to possibly include both species for planning purposes (Figure 4).
2.2.1.2 Habitat modeling for bird priority zones
Within the research project ‘Bird Priority Zones for Species Protection’, the premise was to test a method for bird priority zones approaches that would produce robust results with minimal data and uncertainties in the quality of mapping data for habitat potential [91]. Habitat modeling provides an efficient way to interpolate the distribution and occurrence of target species based on a sample [86, 85, 84]. The method is based on niche theory, which states that a species can only survive if both abiotic and biotic interactions allow positive population growth [92, 86, 93].
For the target species, Red Kite and Osprey, point occurrence data were provided for the state of Brandenburg by the State Office for the Environment (LfU), Brandenburg. To identify further suitable nesting sites in potential habitats, habitat parameters were considered, e.g., forest, grassland, water, wetland. A logistic regression was used to calculate the probability of occurrence for characteristics and constellations of habitat parameters, assuming only two possible values (yes, no). The first five and the first ten most suitable habitats were selected as two options for bird priority zones (Top 5, Top 10). The sites were identified considering the criteria of contiguous area size, habitat quality and suitability for other species. Depending on how much space is available, priority areas can be selected more generously from the pool of potential areas, or focus only on those areas that are suitable several times over. In Havelland-Fläming, there are no suitable areas designated as Top 5 for the Osprey. More detailed information on the habitat model can be found in the Supplemental Materials.
2.2.2 Variations of energy and spatial targets
2.2.2.1 Spatial targets for wind energy
The federal government in Germany has introduced binding targets on the amount of land that must be designated for wind energy in the German states (Länder). These targets must be met by 2023[5]. As a benchmark for this analysis, the targets for wind energy sites in the state of Brandenburg were examined and are referred to as ‘spatial targets’ (compare Table 1). As the states are allowed to adopt spatial targets, which are even more ambitious[6], in addition, we included the targets of the state of Brandenburg. The 2030 energy strategy sets targets for designated wind energy sites (2.0%) [94]. The 2040 energy strategy expands energy targets for the next decade, taking up the nationally defined target of 2.2% for wind energy sites[7] [95]. The different targets allow for the analysis of different time horizons. The spatial targets may include areas that already feature wind turbines (Table 1).
2.2.2.2 Spatial targets for ground-mounted photovoltaics
For ground-mounted PV, there are no spatial targets for a number of designated sites set by federal legislation. Instead, energy production targets are set, which are intended to be achieved on across Germany[8] and include both roof-mounted and ground-mounted PV. Assumptions were made specifically for the distribution of federal targets in the state of Brandenburg. It was assumed that half of the PV targets would be met by roof systems and half by ground systems. We postulated an equal distribution key for the 16 federal states, except for the three city states. For Brandenburg, the calculated energy target must be implemented in the state's four regions
[71]
. We referred to an equal distribution of PV systems, i.e. a quarter of the energy target for Brandenburg would be installed in the Havelland-Fläming region. Based on a power density per area
[97–99]
, the area required to achieve the energy target for PV was then calculated (section 2.2.3). In addition, Brandenburg's PV development targets were added for comparison
[95]
(Table 2).
2.2.3 Calculating capacity density per area
2.2.3.1 Calculations for wind energy
Capacity density assumptions are required to determine how much wind power can be installed per area. Capacity density typically varies due to geographic and topographic conditions such as wind speed, surface roughness (e.g., in forest areas), altitude, and the density of wind turbines in an area
[100–102]
. For this analysis, the average value of 29.3 MW/km2 for Germany is used
[100]
. This value assumes a rotor pitch of 4.5 times the diameter in each direction. However, it must be acknowledged that other capacity densities could potentially be achieved for the case study area. The shape of how wind turbines are arranged to another influences the power density, but this cannot be predicted in detail for each possible site
[100]
.
Furthermore, different power densities can be postulated, depending on the turbine capacity, height and rotor diameter
[103]
. Based on market developments
[104–107]
, a reference turbine with a capacity of 5 MW, a rotor diameter of 149m and a hub height of 150m is assumed for this analysis [such as 108–110].
2.2.3.2 Calculations for ground-mounted photovoltaics
For ground-mounted PV, different capacity densities can occur depending on the system type, i.e., ground-mounted, elevated on agricultural land, or as vertical PV systems [111]. The capacity density depends on factors, such as technological efficiency, position of the sun, slope and shading [97, 98, 112]. Based on a literature analysis of current system data, a power density of 99 MWp/km2 is assumed for ground-mounted PV systems, e.g. installed on grassland [97–99]. In Germany, ground-mounted PV systems are subsidized for certain areas (referred to as ‘EEG areas’)[9]. In the case of agri-PV, the power density is lower at 40 MWp/km2 [97]. Vertical PV modules allow the use of solar radiation from both sides in an east-west direction [113, 111]. These modules can be installed on extensively used grasslands, for example, and have a power density of 35 MW/km2 for ‘Next2Sun’ modules [114].
2.2.4 Generating scenarios and data application
Scenarios should simulate renewable energy allocations using bird priority zones under defined renewable spatial targets. To give an example and allow evaluation, the data input and results are assessed based on the case area (compare section 2.1.1). This exemplary analysis therefore aims to help exploring how to achieve the remaining site deficit for wind energy, i.e. the ‘delta’ of 0.56 % of the case region, while considering bird priority zones (section 2.1.2). Thus, the existing draft spatial planning concept of the case study region is modified in scenarios and used for wind energy according to its planning criteria (cf. section 2.2.4.1 and Appendix A). This approach is exemplary and can also be applied to other case regions by supplementing or adapting the energy targets and the set of criteria. The analysis also investigates the spatial potential for using ground-mounted PV as a complementary approach to achieving the spatial delta (based on a informal potential site analysis for solar energy [115]). The relationship between wind energy and PV for energy target achievement is therefore explored, as well as how bird priority zones may affect the development of both. A detailed and criteria-based analysis similar to the one for wind energy was not carried out, as such an analysis was not conducted in the cases’ regional plan [78], and due to the scale of a similar self-made analysis. However, PV was included here as a complement based on an informal spatial PV analysis [cf. 115]. As land use pressure continues to increase, it is important to balance stakeholder desires for PV in conjunction with wind energy, to avoid separate considerations [116, cf. 117]. This framework would also allow for the inclusion of more detailed criteria-based analyses for PV.
The scenarios represent possible futures to guide decision making
[118, 119]
and simplify contextual complexity
[120]
. For the purposes of this analysis, the possibility of alternative futures is assumed. Future events are not completely predictable, but they are also shown to not be completely chaotic, allowing for possibilities of influence, for example through goals in decision-making processes
[55]
. Scenarios for describing the future, which have an exploratory nature in this analysis, are used to capture 'if-then' interactions
[55, 121]
. Starting from a 'business as usual' (BAU) reference, a leap in time is mapped into the near future
[122]
, such as the usual planning cycles of about 10 years in regional planning
[123]
.
Based on these assumptions, the intuitive logic scenario generation method is identified as appropriate. It focuses on decision-making processes where intuitive ‘gut feelings’ and uncertainty assessments are allowed as well as objective data and information
[55, 124]
. There are different phases in the scenario generation process, such as key criteria identification, key criteria analysis, and scenario development. Participatory evaluation can be done at the end, but is left for further analysis here [55, 56, 118].
2.2.4.1 Key criteria identification
Since the scenarios aim to vary planning criteria for the allocation of wind energy based on the case study, the planning criteria were adapted from the draft regional plan of Havelland-Fläming [125]. The goal was to model the draft regional plan of the case study region for the BAU-scenario, and vary them in scenarios. A PESTEL analysis (Political, Economic, Social, Technological, Environmental, Legal analysis) was used to analyze the planning criteria that were applied
[126–129]
. In addition, the planning criteria were supplemented by the bird priority zone approaches (section 2.2.1). We also addressed further criteria that arose within the region, such as repowering potential, retrofitting, and reduced buffers to settlements
[130–133]
.
In the case of ground-mounted PV, a state and regional informal analysis of possible PV sites is available via a Web-GIS application [115] (cf. section 2.2.4). This analysis covers areas for which subsidies are available for the development of PV (section 2.2.3.2) (‘EEG areas’). Additionally, the possible areas for agri-PV are assessed
[115]
. Since the results can only be viewed online, no detailed criteria-based analysis was conducted, but the availability of sites was included as a correction factor for the PV scenarios (see Supplementary Materials for detailed GIS approach).
2.2.4.2 Key criteria analysis and (geo-)data availability
In a key criteria analysis, only the driving forces among the planning criteria were identified to combine them to derive scenarios
[56]
. It is assumed that criteria with little uncertainty about land use demand can be combined into a single profile throughout the scenario process. Criteria with ‘critical uncertainty’ should be considered in the form of possible scenario profile curves [55]. A matrix of land use impacts and uncertainties was used for this purpose
[56, 55]
. Within the matrix, the criteria were scored in terms of their impact on land availability (y axis). Uncertainty for the policymaker about whether to include or exclude criteria for wind energy is displayed on the x axis (refer to Figure 5). Criteria assignment in the matrix was based on geography in the case study area, and how the criteria were applied:
- Criteria that account for a significant amount of land were assigned a higher land use demand and impact, and vice versa. It can be assumed that those planning criteria that take up the most area will have greater trade-offs with wind energy and PV development.
- For criteria that have a legal basis, the planning scope was considered small and therefore the uncertainty to use this criterion in decision-making also smaller, since there cannot be much scope for action. For example, criteria, which are not subject to balancing, but restrict the area for wind energy due to physical and legal reasons, are nature reserves, military areas, open spaces, installed ground-mounted PV systems, and airports. Additionally, there is a minimum buffer of 1000m to hospitals and 350m to settlement areas [125].
- For criteria that are only considered in planning, i.e. are subject to balancing, the uncertainty was estimated to be higher.
- Criteria that were particularly discussed (in the media) in the case study area were classified as ‘uncertain’, although these criteria can be legally regulated (e.g. landscape protection areas) (refer to Figure 5).
‘Critical uncertainties’ to be tested in the scenarios (see Figure 5) are thus specific criteria with a high land demand in the study area and possibly present a higher uncertainty for decision makers. However, only those criteria for which (geo)data are available for the study area can be run for these scenario profile curves. Even though many (geo)data are already publicly available, missing data for some criteria would still be integral for further analysis. Based on the PESTEL analysis, a data search was conducted to analyze the availability of (geo)data for criteria (Supplementary Materials for geodata sources). Specifically, geodata was not available for species-specific buffers. However, since it is known how much area is occupied by species-specific buffers in the region
[134]
, these buffers were included as a correction factor in the scenarios[10]. In addition, no geospatial data were available for telecommunication channels or low-flying areas of the German Federal Armed Forces
[134]
, which also compete with wind energy
[6]
. On the basis of available data for the Multi-Criteria Scenario Framework, it can be assumed that results closely approximate current regional planning practices.
2.2.4.3 Scenario generation
The planning criteria identified and then combined within the scenarios cover three thematic areas:
- Planning based on the current situation, i.e. according to the third draft regional plan for wind energy and the informal area analysis for ground-mounted PV (Business as usual, BAU)
- Planning variations in bird priority zones
- Planning variations of large-area requiring criteria with fixed bird priority zones
The thresholds within the scenarios refer to the share of land that could be considered for wind energy or PV in theory. They were chosen to have a small influence on the planning criteria, i.e. land uses, and to reach the spatial targets for wind energy and PV (Table 3). A detailed description of the scenarios is provided in the Appendix A.
2.2.4.4 GIS and Excel modeling of scenarios
A Q-GIS graphical model was used to map and analyze the impact on land availability for renewable energies within each scenario using vector data (see Supplemental Materials for data used and detailed model description). The planning concept for wind energy in Havelland-Fläming has been mapped according to the draft regional plan in Q-GIS [78]. Using open access data, the total percentage of space available for wind energy according to the given planning criteria in the BAU scenario was 1.45% of the region's total area. Compared to the identified available space for wind energy of 1.67% in the draft regional plan, this is a slightly lower value [78]. This may be due to the fact that not all necessary data from the regional plan was openly accessible, or other data sets may have been used (see also section 2.2.4.2).
The planning criteria of the BAU scenario were then modified according to the proposed scenarios in Q-GIS (compare Table 3). For example, wind energy development was allowed in a small portion of landscape protection areas. The remaining wind energy area was calculated per scenario. The PV scenarios were based on the informal analysis of possible PV areas of the state of Brandenburg [115]. However, in the absence of geospatial data for GIS, it was assumed in this analysis that there would be an even distribution of potential PV sites across the region (see Supplemental Materials for detailed model description).
An Excel model was then applied to combine all input parameters with the spatial results of the GIS model for the case study (i.e. bird priority zones, spatial targets for wind energy and PV, capacity density per area for each scenario). As a result, it was possible to calculate how much land would remain for wind energy and PV under each scenario, and whether or not the federal spatial energy targets could be met (Table 4, Table 5). The calculations in the Excel model are provided in the Supplemental Materials.
2.2.4.5 GIS-Overlay analysis of planning criteria
A GIS-overlay analysis was conducted to determine whether sufficient land would actually be available for wind energy in each land use, i.e planning criteria, if it were opened to wind energy according to the scenarios. Due to legally mandated criteria that may not be available per se, it is possible that even with good planning intentions, sufficient land cannot be designated for renewable energy. For example, open space areas may overlap with strictly protected areas, such as nature reserves. Therefore, planning criteria that are mandatory have been overlaid with criteria that have a larger area share in the region (see section 2.2.4.2 for mandatory planning criteria) (RPG HF, 2020). The following criteria were identified for overlay analysis using QIS: nature reserves, coniferous forests, nature parks, and open spaces. The results are given in the Appendix B.
For additional discussion, GIS was also used to determine how the region's bird hotspot approaches overlapped with other protected spatial categories, such as Special Protection Areas (SPAs) and nature reserves. The aim was to determine whether the categories were spatially complementary. It was also investigated whether there would be an umbrella effect for other species. The Supplementary Material presents these additional results.
[2] Until 2023, the planning criteria which would not be available for wind energy had to be specified in detail in order not to undertake a preventive planning approach
75, 76
. In 2023, 'positive planning' was introduced to speed up wind energy growth. This sets spatial energy targets for each state avoiding justification of exclusionary criteria. Therefore, we argue that a multi-criteria framework allows for the exploration of potions and criteria for achieving spatial energy goals.
[3] Appendix 1 to §3 (1)
42.
[4] The spatial wind energy target of 2.2 % corresponds to 150.5 km2 for the Havelland-Fläming region (= 15050 ha), 71.
[5] Appendix 1 to §3 (1) 42.
[7] Appendix 1 to §3 (1) 42.
[8] §4 No. 3
96.
[9] such as areas in commercial and industrial zones, areas along highways within a 200-meter corridor, areas converted from economic, traffic, residential, or military use, sealed areas, and landfills, as well as areas in disadvantaged areas, such as low-yield arable land and grassland. Ground-mounted systems on peat land, parking lots and floating PV systems will also be permitted, §37 (1) 96.
[10] The area available for wind energy designation is 0.73 km2 after subtracting all planning criteria. Species-specific buffers occupy 0.42 km2 of this area. This is about 58% of the area (
134.