**1. Introduction**

Rainfall is the most important meteorological and climatological parameter for natural ecosystems and human life on earth, as it affects the enrichment of lakes and underground aquifers, river flow regime, and many natural hazards (floods, drought, landslides, etc.). Accurate knowledge of the spatial and seasonal variations of long-term rainfall time series is required for rural and forest development and planning, sustainable development, as well as infrastructure work scheduling.

The Mediterranean basin has a wide range of climatic conditions [1]. In particular, the rainfall regime in Greece presents a highly irregular behavior, both on spatial and temporal scales, namely in rainfall amount and rainfall distribution [1,2]. It is well accepted that the main physical and physico-geographical factors controlling the spatial distribution of rainfall over Greece are the following: the atmospheric circulation, the mountains in the west and east, the Mediterranean Sea-surface temperature distribution, the dehumidification of the air masses crossing the Aegean Sea, and land and sea interactions [3]. Furthermore, the highest rainfall totals for western Greece were found to be related to the atmospheric circulation associated with the Mediterranean Sea-surface temperature distribution and the complex topography of the region, as imposed by the orography of the Pindus Mountains in northwestern and central Greece and the mountains of Olympus and Crete [4].

Mountainous areas are of great interest, because runoff is generated and supplies lowlands (through catchments) with water. Moreover, the plain areas receive the eroded material deposited by mountainous catchments, due to intense rainfall. Variability is considered particularly higher in a mountainous environment, because the rainfall pattern is influenced by complex terrain conditions [5–7]. The assessment of climate variability is a common issue that should be treated by hydrologists; in particular, the total rainfall in an ungauged site over an area (e.g., catchment) should be evaluated. However, hydrologists face a crucial challenge when it comes to mountainous terrains, since data from only a few meteorological stations are usually available.

To overcome the lack of rainfall data, interpolation methods have been developed over the last few years for rainfall modeling and mapping. These methods are based on the similarity and the topological relationship between nearby sample points and on the value of the variable to be measured [8]. Interpolation can be achieved using simple methods (splines, inverse distance weighting, Thiessen polygons, etc.) or advanced geostatistical methods (e.g., kriging). Geostatistical interpolation has become the most appropriate downscale technique in applied climatology and for areas with complex terrain, since it is based on the spatial variability of the variables of interest and allows the quantification of the estimation uncertainty [9–11].

In recent decades, the interest in climate variability and climate change has augmented. Climate change has emerged as a key issue facing environmental and economic aspects, as it affects floods [12], soil erosion [13], drought phenomena [14], agriculture [15], tourism [16], groundwater aquifers [17], and forest fires [18].

According to IPCC reports [19], the Mediterranean basin is expected to become warmer and drier due to the anthropogenic increase of greenhouse gas (GHG) emissions (CO2, CH4, N2O, and F-gases), until the end of the 21st century [20,21]. Moreover, in Mediterranean regions, future warming is expected to be greater than the global mean, accompanied by a significant decrease in rainfall [22]. Based on the above, researchers are orienting their work to investigate trends in rainfall conditions [23–28] and to estimate future rainfalls [29,30] within Greece. Research results highlighted the decreasing trend of rainfalls recorded from long-term time series analysis, whereas this reduction is expected to be higher in the future, based on regional climate models (RCMs) that have been proposed. Even though much research has been conducted in Greece on trend analysis and spatial mapping of rainfall, only limited research efforts concern mountainous areas with consideration given to long-term time series using a dense network of stations. The identification and recording of seasonal trends can improve water resources management through the selection of appropriate management practices.

The main object of this study was to detect annual and seasonal variation and trends in rainfall time series based on data from rain gauge stations located in mountainous areas. Furthermore, variation and uncertainty in the small-scale rainfall interpolation in mountainous catchments were also evaluated.

#### **2. Materials and Methods**

#### *2.1. Study Area*

The study was conducted over the central Pindus mountain range, in Central Greece. The area is considered highly important from a hydrological point of view because it is located in the mountainous area of two hydrological basins (Pinios and Acheloos), which supply Central Greece with water, and where many hydropower dams have been constructed. For this purpose, a dense network of meteorological stations (compared to other regions in Greece) has been established, in mountainous terrain. The characteristics of the meteorological stations used in this study are given in Table 1.


**Table 1.** Meteorological stations in the study area.

Observations of monthly rainfall totals for a period of 55 years of rainfall (1961–2016) were used from all nine stations of the wider region (see Figure 1). These stations are equipped with pluviometer and Hellmann-type rain gauges (Fuess Meteorologische Instrumente KG, Königs Wusterhausen, Germany) with a precision of 0.1 mm. The data series are complete, that is, they have no missing values. Moreover, the instruments and observing practices were common among all stations used, and they remained the same during this study's research. The double mass method and two parametric statistical tests (Student's *t*-test and chi-squared test) were applied to adjust any heterogeneity of the rainfall data, and the details regarding these methods can be obtained from the WMO [31]. The latter tests demonstrated that the precipitation data were indeed homogeneous and ready to be entered into the subsequent procedures of the study.

**Figure 1.** The study area and location of stations.

These stations are operated by the Ministry of Environment & Energy (Agiofylo, Chrysomilia, Elati, Katafyto, and Malakasi), the Public Power Corporation (Mesochora, Polyneri, and Stournareika) and the University Forest Administration and Management Fund (Pertouli).

The study area is an area of increased importance, because it is located in the mountainous area of two hydrological basins (Pinios and Acheloos), which supply Central Greece with fresh water. The mountainous catchments examined within this study are (1) Klinovitikos, (2) Aspropotamos, (3) Korpos, and (4) Portaikos, as showed in Figure 1. Additionally, the basic morphometrical and hydrographical characteristics are given in Table 2.


**Table 2.** Morphometrical and hydrographical characteristics of the mountainous catchments.

The study area is characterized as mountainous, whereas the relief is rather intense. Regarding geology, the main rocks are flysch and limestones, quite vulnerable to landslides and weathering phenomena. The forest cover is high and distributed to the mountainous catchment as follows: (1) Klinovitikos, 66%; (2) Aspropotamos, 73%; (3) Korpos, 72%; and (4) Portaikos, 44%. The dominant forest species in the study area are *Abies borisii-regis*, *Quercus frainetto*, *Quercus petraea*, *Pinus nigra*, and *Fagus Sylvatica*. Moreover, the study region is of great environmental importance, belonging to the European nature conservation network Natura 2000 according to the criteria of Directive 92/43/EEC.
