A Novel Approach to Assessing Carrying Capacity for Development by Combining Socio-Economic and Environmental Indicators: A Case Study in Greece

Abstract

Carrying capacity for development refers to the maximum level of economic and social growth that an environment can support without experiencing significant degradation or collapse. This concept integrates environmental and socioeconomic dimensions to ensure sustainable development. In the present work, we have combined climate reanalysis data sets with environmental and socioeconomic indicators to develop a novel index, the Carrying Capacity Development Index (CCDI). Thus, the Human Modification Index (HMI) and the Vulnerability to Disasters Index (VDI) were examined as part of the socioeconomic dimension of the CCDI, while selected ERA5 land layers combined with Natura 2000 sites formed the environmental elements. The methodology is demonstrated in Greece, where economic development in sectors such as tourism, transport and energy are threatening fragile Mediterranean ecosystems. Three different weightings were analyzed, giving equal weight to environmental and socioeconomic factors, or giving more weight to either the environment, society or the economy. The results showed that the methodology has a minor sensitivity to weighting and highlighted areas where future development objectives should be focused.

1. Introduction

Nowadays, many countries consider human advancement and destitution mitigation as the benchmarks of development and success. Moreover, human development plays a critical role in increasing the well-being of the people in a country [1], but at the same time it poses various environmental and socioeconomic challenges [2,3,4,5]. Also, some studies have shown that changes in land development have led to some health and environmental injustices. Although the standard solution to poverty is economic growth, there is still evidence that this approach is being undermined by the unprecedented levels of consumption which have resulted in major public health problems like obesity [6]. Moreover, by ignoring the damaging consequences of environmental injustice, organizations’ responses to natural disasters fail to address civilians’ substantial loss of confidence and trust [7].

The concept of carrying capacity for development refers to the maximum level of human activity and population that an environment or system can sustain over the long term without degrading. It incorporates ecological, economic and social dimensions, emphasizing the need to balance development with the sustainable use of resources [2,8].

Sustainable development is a dynamic concept that requires a balance between growth and sustainability. It requires a holistic approach that integrates environmental, economic and social factors to ensure that development can be sustained over the long term without compromising the ability of future generations to meet their needs.

In many areas of the world there is a growing concern about the impacts of development on the human and natural environment. The development of tourism, for example, is known to provide an opportunity for prosperity to various communities, especially mountain ones, where other development options are limited. However, tourism impacts fragile environments, and therefore, management strategies should be evaluated in order to achieve sustainable tourism. Thus, an index, namely the Tourist Carrying Capacity (TCC), has been proposed for the Kashmir Valley, so as to regulate tourist flow in the most popular tourist destinations and minimize the environmental and socioeconomic consequences [9].

The Mediterranean is another popular tourist destination where overtourism has been criticized for its pressure on local communities and heritage. One study investigated the impacts and social carrying capacity of tourism in a Catalonian destination. It found that there is a consensus among the local population on the impacts of tourism on conservation. Furthermore, the study highlighted the issue of there being limited space for residents, which affects various capacity indicators [5].

The urban land carrying capacity is a more specific concept and examines the impacts of urbanization on land resources and tries to quantify an acceptable development threshold that supports sustainable development [2,10,11]. The use of land for the energy sector, e.g., the installation of renewable and conventional energy plants, wind turbines, solar panels and other bioenergy land uses, poses several issues in terms of the sector’s land footprint and the land capacity that is required to sustain the expanding number of energy land uses, along with its feasibility for mitigating climate change [10]. Other works have focused on the wider concept of the carrying capacity of land in general, without focusing on a specific land use type (e.g., urban) or a specific development pattern (e.g., tourism or urban) [8,12], and the paramount impact of land uses on habitat quality has also been documented [13].

Carrying capacity has been linked to sustainable development and it was defined as an operational tool for sustainable development [4]. Previous works have also highlighted the potential of using carrying capacity as a measure of the achievement of the United Nations’ Sustainable Development Goals (SDGs) [3] and the need for scientific initiatives that will support the estimation of the SDGs’ success within the framework of carrying capacity [14].

The accelerated growth of open datasets and the expansion of computational resources have facilitated the development of a range of indices that facilitate informed policy decision making across many sectors of human activities.

In the present work we developed a new index to evaluate carrying capacity at the country level, namely the CCDI, integrating various indicators that describe the modification state of land, the vulnerability to disasters (also including the social vulnerability dimension) and the physical state of examined system in terms of the trends in water resources and vegetation conditions. Policymakers and planners need a simple yet reliable tool that adopts a holistic approach, incorporating ecological footprints and resource use, in order to set adaptive management policies to align development goals with environmental sustainability. We thus provide an operational tool which is easily evaluated and updated for policy makers to enhance sustainable development and society’s well-being.

2. Methods

The computation of CCDI is based on the integration of six layers of information that describe the physical and human-related state of land; thus, the Human Modification Index (HMI), the Vulnerability to Disasters Index (VDI), the Total Water Storage Changes (TWSCs), the Leaf Area Index for High Vegetation (LAIH), the Leaf Area Index for Low Vegetation (LAIL) and the Natura 2000 network of nature protection areas (NAT) were combined in an index and overlay analysis. All the above-mentioned factors describe through their sub-indices or their temporal trend the state of land and natural resources related to human activities and climate change. The CCDI ranges from 0 (least carrying capacity for development) to 1 (highest carrying capacity for development). Thus, all examined indicators were normalized to the range of 0 to 1, with 0 corresponding to areas that are heavily modified and carrying capacity is exhausted and 1 to natural systems that are not altered and thus their carrying capacity for development is maximum. The normalization approach proposed in [15] was adopted. Thus, two mathematical formulas were used to normalize each indicator based on the relationship between CCDI and the examined indicator as described in

Table 1. The six indicators used to build the CCDI.

Indicator	CCDI
Human Modification Index (HMI)	(−)
Vulnerability to Disasters Index (VDI)	(−)
Total Water Storage Changes (TWSCs)	(+)
Leaf Area Index for High Vegetation (LAIH)	(+)
Leaf Area Index for Low Vegetation (LAIL)	(+)
Natura 2000 (NAT)	(+)

. For example, when higher indicator values correspond to higher suitability for development and thus higher CCDI, the rescaling method is described by Equation (1):

$$Indicator Score=\frac{x-Min x}{Max x-Min x}$$

(1)

Otherwise, if higher indicator values correspond to lower suitability, then the rescaling method is described by Equation (2):

$$Indicator Score=1-\frac{x-Min x}{Max x-Min x}$$

(2)

In Equations (1) and (2), x is the actual value of each pixel that will be rescaled; Min x is the minimum value of the raster layer; and Max x is the maximum value of the raster layer.

Initial and normalized layers of examined indices from Equations (1) and (2) are presented in Figure 1 and Figure 2, respectively.

Although there are various sophisticated approaches to combining and weighting the different indicators under consideration, such as the entropy approach [2], the process needs to be kept as simple and understandable as possible for policymakers to have an operational tool. Eklund et al. [15] proposed the identification of various dimensions and the equal weighting of indicators within each dimension. The present work identifies two dimensions that interact in determining the carrying capacity for development, namely the environmental and the socioeconomic dimensions. Thus, HMI and VDI form the socioeconomic dimension, while TWSCs, LAIH, LAIL and NAT make up the environmental dimension. The weights assigned to the various examined indicators are shown in

Table 2. Weights distribution across the examined indicators.

Indicator	Socioeconomic	Environmental	Indicator Average Weight
HMI	0.50		0.25
VDI	0.50		0.25
TWSC		0.25	0.125
LAIH		0.25	0.125
LAIL		0.25	0.125
NAT		0.25	0.125
Sum of weights	1	1

. The aggregation of the aforementioned two dimensions is contingent upon the degree to which environmental and socioeconomic sustainability are considered equally important, or a skew towards either of those factors is desired. In this work three different weightings of environmental and socioeconomic dimensions are presented. The first approach assigns equal weight to both dimensions, the second is skewed towards the environmental dimension, and the third is skewed in favor of the socioeconomic dimension as shown on

Table 3. Weight distribution across the environmental and socioeconomic dimensions.

	Socioeconomic	Environmental	Sum
Equal weights	0.50	0.50	1
Socioeconomic skewed	0.70	0.30	1
Environmental skewed	0.30	0.70	1

. Results are presented in the form of five equal range CCDI categories. Thus, CCDI ranging from 0 to ≤0.20 is assigned to LOW category. CCDI between 0.20 to ≤0.40 is assigned to the MODERATE LOW category. MODERATE category corresponds to CCDI from 0.40 to ≤0.60, MODERATE HIGH is assigned to the range from 0.60 to ≤0.80 and HIGH CCDI is related to the range from 0.80 to ≤1.00.

3. Description of Datasets

Four data sources that provided six indicators have been incorporated in the present work, as shown in

Table 4. Summary of indicators examined in this study.

Indicator	Technique	Parameter	Spatial Resolution	References
HMI	Remote sensing/survey/global datasets	Human Modification 2017	1 km	[4,13]
VDI	Index aggregation at different dimensions and at various scales. Indices obtained from pan European datasets from Eurostat, World Bank, UNESCO, Worldwide Governance Indicators, European Environment Agency, University of Gothenburg, World Resources Institute and Copernicus (CORINE)	Vulnerability to Disasters index 2022	Vector layer at EU NUTS3 level, converted to 9 km resolution raster	[3,12]
TWSCs (ERA5-Land)	Hydrologic modeling and data assimilation of remote sensing products	Estimation from monthly P, ET and R provided in the ERA5-land dataset	9 km	[14,15]
LAIL (ERA5-Land)	Hydrologic modeling and data assimilation of remote sensing products	Leaf area index low vegetation	9 km
LAIH (ERA5-Land)	Hydrologic modeling and data assimilation of remote sensing product	Leaf area index high vegetation	9 km
NAT	Policy instrument, Compilation of the data submitted by the Member States of the European Union	Natura 2000 sites	Vector layer converted to 9 km resolution raster	[8]

. Since these products are well documented in previous publications (see column 5 from Table 4), we will summarize only their key features and processing prior to their introduction into the computational approach proposed here.

3.1. Human Modification Index (HMI)

The HMI dataset used in the evaluation of the CCDI is presented in [16]. In their study, Theobald et al. [16] produced detailed global datasets for several years, i.e., 1990, 2000, 2010 and 2015, using Google Earth Engine (GEE) platform [17] to assess the spatial and temporal patterns of land modification on the Earth’s surface. In addition, a new estimate of the contemporary, i.e., 2017, land modification was provided, and it was found that 14.6% of land has been altered, which is more than the size of Russia. The produced datasets are very detailed (with 0.09 km² resolution), cover a long period (1990–2015) up to 2017, include a lot of different factors that can cause change (11 change stressors, 14 current stressors for 2017), are reliable (using a validated approach and taking into account classification errors and uncertainty, and have been tested for their accuracy.

In the current study, we used the most recent dataset i.e., that from 2017 (assuming that it represents the contemporary state of land modification), which was produced by aggregating the 14 stressors, and is available online at [18] (access date 10 March 2024).

3.2. Vulnerability to Disasters Index (VDI)

The VDI describes the disaster resilience of communities based on their social and economic situation. Several studies have demonstrated that, when a disaster happens, socially vulnerable people are more likely to suffer and less likely to recover. It is essential to identify and address social vulnerability, in order to reduce human suffering and economic costs after a disaster [19]. Vulnerability is a risk component considered in disaster risk evaluation, in conjunction with exposure and hazard [15]. As a result, as vulnerability is one of the main drivers of risk, it is considered to be necessary to monitor and measure it [15,20].

According to Eklund et al. [15], since natural disasters will not stop happening, the best way to reduce the risk of disasters is to make people less vulnerable to their effects. Thus, in their assessment, they introduced an indicator for measuring how vulnerable countries in Europe are. VDI has four dimensions: social, economic, political and environmental. It shows how vulnerable a system is to disasters at different levels of government (national, regional and local). The VDI ranges from 0 to 10, which includes the four different aspects in a complex way. The information included in their analysis uses the data and framework for vulnerability from the Disaster Risk Management Knowledge Centre (DRMKC) Risk Data Hub (RDH) to examine the risk of disasters resulting from various hazards in Europe.

The argument put forth in the present work is that, as the value of the VDI in a specified area increases, the likelihood of that area supporting sustainable development is reduced, thereby reducing its carrying capacity. Thus, VDI is incorporated into the present work for Greece at the most detailed scale, i.e., the NUTS3 administrative level, for the latest available assessment year, i.e., 2022, which is available online at [21] (accessed on 10 March 2024).

3.3. ERA5-Land

Water availability is a key component of economic development, while water scarcity is posing a serious threat to global economic growth, health and security [22,23]. In order to describe the availability of water resources, the rate of total water storage changes was estimated by differentiating the water budget equation provided in Equation (3) [24]:

$$TWSC=P-ET-R$$

(3)

where P stands for precipitation, ET corresponds to evapotranspiration and R is runoff. All of the parameters in Equation (3) are expressed in mm units. In order to define the long-term trends in the water budget as expressed by TWSC, a derivative of Equation (3) over time was estimated:

$$\frac{∆TWSC}{∆t}=\frac{∆P}{∆t}-\frac{∆ET}{∆t}-\frac{∆R}{∆t}$$

(4)

The terms on the right-hand side of Equation (4) were estimated at the pixel level by least squares fitting [25] for the time series of P, ET and R, from 2000 to 2023. P is the monthly precipitation, ET is the monthly evaporation and R is the sum of surface and subsurface runoff provided in the ERA5-land dataset [26], which was accessed and processed within the GEE platform. The monthly aggregates of the P, ET and R of the ERA5-Land dataset were used. ERA5-Land is the result of an initiative of the European Centre for Medium-Range Weather Forecasts (ECMWF) within the Copernicus Climate Change Service which aims to update the land component of ERA5 climate reanalysis, providing improved an spatial resolution of 9 km compared to the 31 km resolution of ERA5. The model information is combined with observations aiming to describe the water and energy cycles accurately over land, providing hourly data from 1950 until today. In the present work, the ECMWF ERA5 Land monthly aggregates of the hourly assets of ERA5-Land [26,27] were used, which is a readily available dataset on the GEE platform. Time series of the following ERA5-Land bands were used: (a) total_precipitation_sum corresponding to P in Equations (3) and (4), (b) total_evaporation_sum which describes the ET term and (c) the runoff_sum which is the sum of both surface and sub-surface runoff describing the R term in Formulas (3) and (4). The trend shown in the time series of the above-described variables from 2000 to 2023 was estimated using least square fitting with a Java Script code (available at the Data and Code availability section), and the trend of TWSC (left hand side of Equation (4)) was obtained. A positive trend for TWSC corresponds to an increasing water budget, and an increased potential water availability for various uses.

In an analogous way, the trends in vegetation were estimated using two additional layers of ERA5-Land, i.e., (a) the leaf_area_index_high_vegetation which describes the area fraction of ground covered by high vegetation type, i.e., forest and tree areas, and (b) the leaf_area_index_low_vegetation which describes the area fraction of ground covered by low vegetation types, e.g., shrubs. The same approach to trend estimation was applied for vegetation trends and two additional layers were produced, i.e., trends in high vegetations types (LAIH) and trends in low vegetation types (LAIL).

3.4. Environmental Protection

Natura 2000 is a policy instrument aimed towards protecting biodiversity in the European Union. It is based on the 1979 Birds Directive and the 1992 Habitats Directive and it was operationally implemented through the European Natura 2000 database which comprises a compilation of information provided from the Member States of the European Union and is usually updated once a year [28].

In 1992, the European Union agreed to the habitat directive for the Natura 2000 program (Council of the European Communities 1992). This directive helps to protect (semi-) natural habitats and certain plants and animals in Europe. The countries involved have agreed to create places to protect the specific habitats and species listed in the directive. Concurring with the directive, any negative effect on these ensured resources must be dodged and their condition may not be subject to any debasement. In order to guarantee the successful completion of this objective, it is necessary to monitor specific locations and to provide an update on their condition on a regular basis [29].

By the end of 2021, about 26% of EU land was protected. Of this amount, 18.6% concerned Natura 2000 sites and 7.4% other special treaties designated by individual countries [28] (accessed on 20 April 2024).

In the context of this study, Natura 2000 ver. 2022 was downloaded for Greece from [30] (accessed on 20 April 2024). The dataset has 202 Special Protection Areas (SPAs) and 241 Sites of Community Importance (SCI), two of which are still under consideration.

Although the implementation of Natura 2000 does not aim to stop economic activities, it requires certain rules for the protection of biodiversity [8], and there have been several challenges identified regarding the competing interests of various stakeholders in the designated areas [20]. In the work of [31], landowners, natural resource-based industries and also local municipalities are reported to have frequently objected to the Natura 2000 implementation. We therefore argue that although the carrying capacity of Natura 2000 sites is maintained, their potential to support major economic activities (e.g., massive tourism investment, large extractive industries) is reduced. This aspect is reflected in the weight assignment process described in the previous sections.

3.5. Links to SDGs

The CCDI incorporates many indices and sub-indices that are related to the achievement of the 17 SDGs. The contribution of each index is not necessarily positive regarding the achievement of specific SDGs. For example, the higher the VDI and HMI the lower the achievement of SDGs, whereas ERA5-Land contributes mostly positively (positive changes in TWSC and vegetation indices contribute positively to the achievement of SDGs). However, the information provided by the HMI, VDI, Natura 2000 and ERA5-Land contributes positively to SDG 17, because they are all coordinated initiatives aimed towards the achievement of environmental and socioeconomic welfare. We also believe that the CCDI will contribute towards the SDGs since it aims to support economic development and associated policies that will respect environmental and social welfare. Table S1 in the Supplementary Materials presents the relationship between each examined indicator and their sub-indices and specific SDGs.

4. Results

The results of the CCDI computation based on the three different weightings of environmental and socioeconomic dimensions are shown in Figure 3, Figure 4 and Figure 5. The CCDI classes are defined by the CCDI mean and standard deviation values in each one of the three examined weightings.

Figure 3 presents the CCDI results assigning equal weight to both dimensions, and it is considered a balanced approach. In Figure 3, it is clearly shown that major urban centers as well as the majority of Aegean islands present a CCDI index ≤ 0.60, indicating a low potential for further development, although there are still islands, as well as less populated cities, with CCDI index values ranging from 0.40 to 0.80, which can be perceived as areas with moderate potential for further development. On the contrary, very small urban centers with sparse populations as well as the mountainous regions of Greece demonstrate CCDI values >0.60, indicating the availability of resources for further development. From Figure 3, it can be concluded that the CCDI is relatively high in mountainous areas, while in urban agglomerations it is relatively low, showing a trend of gradual increase from the SE to the NW. The reason for this is that mountain areas are rich in natural resources and less affected by human activities, and therefore appear to be more capable of future development. The Aegean islands are a special case where specific measures to promote sustainable development should be examined, as they are limited in area and natural resources and face the challenge of a significant amount of tourism development that threatens to alter their natural character.

Figure 4 shows the CCDI estimated by skewing weights in favor of the socioeconomic dimension, i.e., weights of 0.70 and 0.30 assigned to the socioeconomic and environmental dimension, respectively. In general, compared to Figure 3, almost the same spatial pattern is observed, with the islands and major urban centers demonstrating low CCDI values and, consequently, a limited potential for further development. As in Figure 3, almost all coastlines are categorized in the second as well as in the middle CCDI classes, indicating that coastal areas are already economically developed and heavily impacted, and consequently they cannot accommodate additional carrying capacity for development. In contrast to Figure 3, in Figure 4 it is observed that more mountainous areas as well as some cities in rural areas have CCDI values >0.80, which indicates that, based on this weighting scenario, there is a high potential for further development of those areas.

Figure 5 shows the CCDI results when the environmental dimension is assigned higher weight, i.e., weights 0.70 and 0.30 assigned to the environmental and socioeconomic dimensions, respectively, which is considered as the environmentally sustainable approach. The same spatial pattern as in Figure 3 and Figure 4 is observed, and same CCDI classes are found. This time, however, only a small area in northeastern Greece has a CCDI value >0.80 and most of the Aegean islands have CCDI values from 0.20 to 0.40, which indicates a lower potential for further development of those areas in this case. Thus, it seems that the environmental conditions in those parts of the country cannot support additional economic activities without jeopardizing the natural resources and environmental quality. Regarding the impact of the individual indicators on the CCDI, the low values of TWSC and LAIH seem to control the CCDI values in SE Greece, along with the Aegean islands, highlighting that various natural resources in those areas are depleted.

Conversely, there are regions in the north and north-west of the country with CCDI values above 0.60, where economic activity is less prevalent and the potential for balanced development is evident. This should be the country’s future aim, as an alternative to the lower development potential of the south-eastern regions.

With regard to the modification of results in terms of CCDI categories and associated areas,

Table 5. Areas occupied by various CCDI categories according to the three examined weightings.

			Total Area (km2)
CCDI Category	CCDI	Equal Weights	Socioeconomic Skewing	Environmental Skewing
LOW	0 ≤ 0.20	81	81	81
MODERATE LOW	0.20 ≤ 0.40	14,606	17,108	12,831
MODERATE	0.40 ≤ 0.60	77,873	69,722	81,343
MODERATE HIGH	0.60 ≤ 0.80	40,268	41,317	40,349
HIGH	0.80 ≤ 1.00	2421	7021	646

presents the areas classified into various CCDI categories under the three weighting approaches. Thus, it can be concluded that the “Low” CCDI class is not altered by skewing weights towards either dimension. An increase in the number of areas classified as belonging to the “Moderate Low” CCDI category is observed when the distribution of weights is skewed towards the socioeconomic dimension. Conversely, in the “Moderate” CCDI category, the opposite is observed. In the case of the “Moderate High” category, the performance of the equal weights and environmentally skewed weights is almost identical. However, the socioeconomic weight skewing results in more areas being classified in this category, although the change is not substantial. The “High” CCDI category is the one that has undergone the most alteration. This is due to the socioeconomic skewing, which has resulted in almost triple the number of areas being classified in this category compared to the equal-weight scenario. The application of weights that skew towards the environmental dimension has the effect of almost eliminating this category. It is important to note, however, that even in the most favorable scenario (socioeconomic skewing), only a small area of approximately 5.5% of the country falls within this category. The preceding results demonstrate that all three weighting approaches yield comparable results, with the “Moderate” category encompassing the greatest number of areas, followed by the “Moderate High” category, the “Moderate Low” category and the “High” category. The two most extreme CCDI categories collectively account for a relatively minor portion of the country, according to the three examined cases. Thus, the methodology presented herein proves to have only a minor sensitivity to weighting of different dimensions, both in areas of various classes and also in the spatial pattern of the CCDI, providing a consistent approach for the computation of the CCDI.

5. Discussion

Compared to other approaches to carrying capacity estimation presented in previous works, our approach is simple yet robust as it incorporates many different aspects of development through the examined indicators and their sub-indices. Guo et al. [22] predicted the population size and economic growth that Yellow River Basin could support in 2030, based on the amount of freshwater availability, under various development scenarios in China. In their study, they used a water resource carrying capacity (WRCC) prediction model, without examining other indicators that may favor or inhibit development. In our work, we introduced various indicators that reflect both the socioeconomic and environmental aspects of development. For example, the HMI incorporates stressors related to various human activities like urban/built-up area expansion, agriculture, energy production and mining, transportation and service corridors, wood harvesting, river damming, human intrusions and air pollution. On the other hand, the VDI examines the resilience of various areas to disasters using the social, economic, political and environmental dimensions at national, regional and local scales. The trend of water resources availability along with vegetation trends were examined through the ERA5-Land dataset, while environmental protection policies were introduced in the computational process through the Natura 2000 protected sites.

Xu and Yang [8] evaluated the land resource carrying capacity (LRCC) in Chongqing city in China They produced a detailed assessment method based on entropy weight-cloud similarity. The methodology was validated using the asphalt pavement experiment as a sample for empirical analysis and it was then applied to Chongqing from 2011 to 2020, incorporating 12 layers of information related to water and soil and ecological environment, as well as the state of social, cultural, economic and technological systems. In our work we developed the CCDI based on readily available and regularly updated global or pan European datasets, also incorporating the vulnerability to disasters aspect, which has not been previously used in other analogous assessments.

Shen et al. [10] proposed a theoretical perspective method named the urban resources environment carrying capacity with the load-and-carrier (URECC-LC). In this method, the urban resources environment is considered as a framework, and the novel concepts of urban loads (UL) and urban carriers (UC) are introduced. They demonstrated how urban water resources are affected in the cities of Beijing, Tianjin, Shanghai and Chongqing in China. The demonstration indicated that the URECC-LC method works well for evaluating URECC; however, it provides a theoretical basis for studying urban resources environments only. Although the approach is interesting, it is specifically designed for urban environments while our approach is better suited to the regional and country levels, looking at all types of human development.

Martín et al. [32] calculated an indicator of social and economic development using the Distance Method of Pena (DP₂) method to measure the disparities in regions of southern European countries in 2006. The DP₂ distance method measures and compares different social indicators across different places and times. In their study, they used an index incorporated from the per capita income, socioeconomic components (i.e., health, employment, education, technological, scientific development) and infrastructure provision. They tried to examine whether the Europe Union regions included in the Convergence Objective for the 2007–2013 program period achieved lower levels of development than the regions no longer included in, by building up a regional classification based on the esteem yielded by DP₂ factors. In summary, they proved that several of the regions included in the Convergence Objective presented a lower level of socioeconomic development than those that misplaced this consideration between 2007 and 2013 as well as the inverse.

Anselmi et al. [33] conducted a study based on multicriteria decision analysis (MCDA) to compare how 27 European countries performed economically, environmentally and socially between 2018 and 2020 using 58 different factors. The MCDA analysis makes it possible to calculate a Sustainable Development Goals (SDGs) index that is positive Sweden, giving simultaneously positive results for Denmark, Netherlands and Finland. The total value is analyzed from the three dimensions of sustainability (economic, environmental and social) in which, once more, Sweden exceeds expectations and has a position of authority in two macro-goals such as energy and security. On the other hand, Denmark has the upper hand when it comes to innovative futures, and the Netherlands does when it comes to the waste cycle. The main implication of their work confirms that Europe is making uneven progress towards the SDG targets, and therefore there is a need to define an agenda that sees greater cooperation between multiple countries.

Just as the European Union was found to be growing unevenly, this study has found that in Greece there are different and uneven growth trends between the different regions. Limitations of the present work are related to the inherent differences in examined indicators. Thus, although the HMI is available at high spatial resolution, ERA5-Land is produced at the 9 km spatial resolution, thus restricting the spatial scale of the analysis to this resolution. This might prove to be quite restrictive in the case of local scale analysis. At the country level, however, the analysis provided enough detail for policy makers to focus their development plans. In terms of temporal resolution, ERA5-Land has a better temporal resolution at the monthly time step used here compared to the VDI, which is updated annually, and the HMI with its latest update in 2017.

6. Conclusions

In the present work a new index for measuring the carrying capacity of areas related to development, namely the CCDI, is presented and its efficacy is demonstrated in Greece. Six readily available indicators related to the economy, society and the environment were incorporated ito the overlay and index computational process. Three different weight assignment scenarios were examined, either balancing the environmental and socioeconomic weights, or skewing them either towards the environmental or the socioeconomic dimensions. All three weightings resulted in comparable results in terms of the CCDI, indicating the consistency of the approach. The ERA5-Land trends in terms of TWSCs and vegetation productivity and the VDI are the factors the shape the spatial pattern of the CCDI, irrespective of the assigned weights. Based on the results, the northwestern and northern regions of Greece demonstrate capacities for further development and have the highest CCDI values. On the contrary, most of the Greek islands seem to be facing environmental and socioeconomic challenges and further development should be examined strictly within the framework of the sustainability of the resources and the fragile ecosystems in those areas.