An Evolution of Climate in Greece by the Johansson, Kerner and Pinna Indices ^†

Abstract

To analyze the climatic evolution of Greece, the ERA5-Land dataset, providing high spatial (~9 km) and temporal (daily) resolution, was utilized. Since the dataset comes from reanalysis (surface, satellite, and radiosonde measurements assimilated with the forecast model), it is ideal for assessing atmospheric variations. Input parameters were used to calculate the Johansson, Kerner, and Pinna climatic indices during the 1964–1993 and 1994–2023 periods. Along with the spatial distribution of the indices for each time interval, the spatial statistics related to the phytogeographical regions of Greece were also demonstrated. The results indicate a clear shift in the climate’s characterization by more “Continental” and more “Moderate Dry” classes, mostly in inland parts of the country. Outcomes involving the mapping material along with spatial statistics data may shed light on the spatial and temporal changes in the climate over the Greek peninsula.

1. Introduction

The Mediterranean basin has been identified as a hotspot for climate change (CC) [1,2]. As such, the Greek peninsula, located in the southeast of the European continent, faces increasingly adverse climatic conditions [3,4]. Rising temperatures, especially during the spring and summer months [5,6], and declining trends in precipitation days [7,8] with an increase in extreme precipitation events [9,10] have been well documented in the past 50 years. These shifts have contributed to increases in aridity and heatwave indices [4,11,12]. Future projections show further surface temperature increases and precipitation decreases in the near (2050) and far future (2100) [13].

The climate indices (e.g., Thornthwaite’s Potential Evapotranspiration [14], Growing Degree Days, de Martonne Index [15]) are exploited for monitoring the shifts in climate and determining the impact of these changes on the agricultural and natural environments. The climate indices are mathematical equations incorporating climatic parameters (e.g., temperature, precipitation), and aid in documenting the spatiotemporal changes in climatic parameters and interpreting their influence in specific scientific fields such as bioclimatology, agroclimatology, and hydrology [2,15].

The Johansson Continentality Index (JCI) measures regional temperature variability, the Kerner Oceanity Index (KOI) reflects oceanic influence on climate, and the Pinna Combinative Index (PINNA) classifies tropical and subtropical climates using annual and driest month precipitation and temperature [2,16,17].

Although climate research has been prominent in the Mediterranean region, the JCI, KOI, and PINNA indices have not been extensively examined in Greece. A robust analysis of the three indices was conducted for Northern Greece [16], Turkey [17], and the Emilia-Romagna region in Italy [18], using surface data from ground stations. Baltas (2006) found that JCI and KOI classified most Northern Greece stations as “Continental,” while PINNA results aligned with the de Martonne Index, indicating a semi-dry Mediterranean climate [16].

Effects from climate change on the agricultural sector may include reduced crop yields, increased irrigation needs, and accelerated plant growth phases [19]. In the natural environment, these changes disrupt species distribution, reduce biodiversity, and raise forest fire frequency, heightening erosion and flash flood risks [20].

As such, the present survey aims to capture the spatiotemporal climatic shifts in the climate in Greece to assess its effects on the agro-ecosystems.

Incorporating phytogeographical frameworks into climate research may prove useful in estimating and better understanding the direct and indirect effects of climate change on plant distribution and behavior [4,21].

2. Materials and Methods

2.1. Study Area

The study area for this survey is Greece (38.85° N 24.4° E), located in the southern part of the Balkan region, with a surface of 131.957 km². Greece is divided into 13 phytogeographical regions (Figure 1).

2.2. Data and Methods

The ERA5 Land Reanalysis Dataset (~9 km) was used for the present study. The reanalysis of data consists of processing historical data from ground stations through Numerical Weather Prediction Models (NWPMs), producing a high-resolution raster dataset. This dataset has been found to have a high correlation coefficient (0.87) with ground station analyses [22].

Table 1. Climate indices, values, and classes.

Index	Value	Class
JCI	>20	Marine
	[20,32)	Moderate Marine
	[32–34]	Transitional
	>34	Continental
KOI	>20	Hyper-Oceanic
	(10,20]	Oceanic
	(0,10]	Sub-Continental
PINNA	>40	Wet
	(30,40]	Moderate Wet
	(20,30]	Moderate Dry
	(10,20]	Semi-Dry
	(2.5,10]	Dry

presents the values and classes of the three indices exploited. JCI is used to assess the influence that continentality has on regional climate based on annual temperature range and the latitude of an area [2]. KOI refers to the oceanity based on the mean air temperature of October, the mean air temperature of April, and the annual temperature range [2,17]. Lastly, PINNA expresses the seasonal and regional water needs, incorporating the annual precipitation, the precipitation of the driest month, the annual mean air temperature, and the mean air temperature of the driest month of the year [2,16].

Hourly air temperature and precipitation data for the two intervals 1964–1993 (p1) and 1994–2023 (p2) were procured from the ERA5 Land Reanalysis using Python v3.12. scripts. The data was analyzed and categorized based on the indices’ thresholds to produce the mean, max, min, and median values for each phytogeographical region, which were then loaded into R v4.5.0 using the Tidyverse v2.0. and terra v1.7. packages to produce the relevant maps for each index and interval [23,24].

3. Results and Discussion

Figure 2 highlights the climatic evolution in Greece as described by the JCI. A clear shift towards “Continental” and “Transitional” classes is evident, with the most prominent changes being denoted in NPi (+26.74% “Continental”), NE (+19.36% “Continental”), Pe (+13.72% “Continental”), Ste (+16.34% “Transitional”), and SPi (+14.26% “Transitional”). Notably, the KiK, Nae, and IoI regions remain approximately unchanged.

The change in climate based on the KOI is depicted in Figure 3. Most of the country demonstrates relatively stable climatic conditions, with shifts referring mostly to a decrease in the “Hyper-Oceanic” class on the west coastline and to an increase in the same class on the eastern part. A few notable shifts are presented for the NE (−13.42% “Sub-Continental”), NC (+4.56% “Sub-Continental”), Pe (+6.51% “Hyper Oceanic”), and the WAe (+35.57% “Hyper-Oceanic”). Moreover, the KiK, KK, Nae, and IoI regions exhibit approximately unmodified climate footprints.

Figure 4 displays the shifts as interpreted by the PINNA. It is evident that drier classes occur from the north towards the south and from the west towards the east. Major shifts towards the “Semi-Dry” class are demonstrated for the EC (+19.74%), while other notable changes for the same class occur in the NE (+18.1%) and NC (+10.48%) regions. Smaller-scale spatial expansions of the “Dry” class are indicated for EAe (+7.08%) and WAe (+6.16%). Phytogeographical regions representing the islands (KiK, KK, IoI, Nae) exhibit small to no changes once again.

4. Conclusions

The outcomes drawn from the present survey can be summarized as follows:

• Based on the JCI, Greece is mostly classified as having a “Moderate Marine” climate; however, clear shifts towards the “Transitional” and “Continental” classes are found mostly on the northern and central areas of the country.
• Regarding the KOI, most of the country is classified as “Oceanic”. The “Hyper-Oceanic” category, predominantly describing the western coastline, demonstrated a spatial decline over the western regions in p2, while the eastern coastline displays a spatiotemporal increase in the same class.
• Based on the PINNA classification, Greece is characterized as mostly “Moderate Dry”, while showcasing all classes of the index. In p2, the “Semi-Dry” and “Dry” classes are spatially expanded, while the “Moderate Wet” and “Wet” classes are restricted.
• Overall, the phytogeographical regions that seem to have faced increased shifts towards drier classes are detected in the northern (NC, NPi, NC) and central (EC, StE, Wae) parts of the country. The climate of most island phytogeographical regions (IoI, KK, KiK) remains practically unchanged.
• The outcomes of the present research illustrate a regional increasing xerothermic trend, which is in agreement with similar studies on historical data [25,26,27] as well as future projection surveys [2,3,4,11].

The methodology exploited in this survey could be further developed by incorporating climate emission scenarios, shedding light on the forthcoming climatic shifts on a regional and country-wide basis and providing insight into their impacts on the agricultural and natural environments. Also the results from similar investigations may aid in developing climate change adaptation measures aimed at the protection of natural resources and by proxy, the socio-economic sustainability and development.