Hail Measurement Characteristics in Central Macedonia—Spatial and Temporal Distribution ^†

Abstract

The objective of this study is the statistical analysis of a six-year period (2019–2024) hit hailpad dataset over the area of Central Macedonia to define the temporal and spatial distributions of selected hail parameters based on hailpad network data. The maximum number of hail days was recorded in June, while the maximum number of hit hailpads was recorded in April and the secondary maximum was registered in June. During the spring months, hailfalls contain hail small in size, while during the summer months, larger sizes of hail are recorded more frequently. The region most frequently affected by hail is the northwest-west part of the area, and locally, the southern part. Pea- and grape-size hail is distributed almost evenly over the whole protection area, but walnut-size hail is mainly recorded over a rather narrow strip that crosses the entire area from northwest to southeast in its central part.

1. Introduction

Hail is a meteorological phenomenon that is difficult to measure and forecast. This is due, in part, to the complexity of the microphysical processes taking place within the cloud during its creation, as well as the variability of its appearance in space and time.

The Hellenic Agricultural Insurance Organization (ELGA) implements the Greek National Hail Suppression Program (GNHSP) [1]. To assess the crop damage caused by hail, as well as to evaluate the effect of the GNHSP in reducing the size of hail, a hailpad network has operated in the area of Central Macedonia since 1984 (Figure 1b), consisting of 157 hailpads in the recent years. Hailpads are the most widely used instruments in the world for recording hail and have been used in many hail studies in Greece and in the rest of Europe (i.e., [2,3,4,5,6,7,8]). Each hailpad carries an ID number and a position defined by the longitude and latitude coordinates of its location in the project area. The average area corresponding to each hailpad is 17.5 km².

For the present research, six hail suppression seasons between 20 March and 30 September 2019–2024 were studied (the start was 15 April 2020). The objective of this study is the statistical analysis of the hailpad dataset from over the area of Central Macedonia to define temporal and spatial distributions of the selected hail parameters.

2. Materials and Methods

A hail day is defined as a day when at least one hailpad with hail dents, namely a hit hailpad, is recorded. A hailpad acts as a witness for hail on the ground, and a specific procedure is established for its collection, storage and evaluation. Each hit hailpad corresponds to a single hail day, with the assumption that it was affected by only one storm.

A calibration process takes place to match the diameter of the dent of a hailstone to its real diameter [9]. Hit hailpads undergo a digital analysis using Image-Pro^® Plus software version 5.1 [10], with some parameters extracted directly and others after calculations. The directly extracted parameters are the minimum and maximum diameter of each dent and the number of dents per hailpad. The calculated parameters are the maximum diameter, mass, momentum and kinetic energy of a hailstone. The kinetic energy of each hailstone is measured in Joules reduced to the unit area (J/m²). The kinetic energy of a hailpad is defined as the sum of the kinetic energies of all hailstones in Joules (J).

A classification of the hailstones into size classes is also made according to the calculated maximum diameter of each hailstone. The classes start from 0.5 cm (the minimum diameter of precipitated ice considered as hailstone). There are three classes: pea (0.5–1.2 cm), grape (1.2–2.0 cm) and walnut (2.0–3.2 cm), each based on the directly extracted diameter of each hailstone.

Each hailstone is classified into one of the three categories above. The largest hailstone on a hailpad determines the category the hailpad is classified in, namely as a pea-size, grape-size or walnut-size hailpad. The hailpad with the largest classification category on a given day determines the corresponding classification of the hail day, as a pea-size, grape-size or walnut-size hail day. For the whole study period, a file is created which includes the number of hail days, the number and location of hailpads with dents per day, the number and size of dents per hailpad, the number of hailstones per hailpad, the maximum diameter of each hailstone, the kinetic energy of each hailstone, and the kinetic energy of each hailpad. The cumulative kinetic energy of a hailpad is consequently defined as the sum of all the kinetic energies in the specific hailpad during a determined time period (month, year or a different period).

3. Results

An interannual and intramonthly analysis of hail days and the number of hit hailpads was performed to record and better understand the hailfall intensity for the preceding six years. For the study period, an average of 19 hail days per year was found, with an average of 5.8 hit hailpads per hail day. The annual distribution of hail days indicates that the maximum number of hail days was recorded in 2019 (23 days), followed by 2023 (22 days), while the minimum occurred in 2020, with 13 days. With the average value of 110 hit hailpads per year, the maximum value of the parameter was recorded in 2023, with 141 hailpads, while the minimum was recorded in 2024, with 65 hailpads.

An analysis of the number of hit hailpads per day showed that, for 37% of the total number of 114 hail days, one to two hit hailpads were recorded. Also, in 75% of the days, one to six hailpads were impacted. This means that when hail occurs, it does not affect a very large surface in the protection area. Days with the number of hit hailpads over 10 were rare, amounting to 16% of the total number. This outcome is consistent with the results in [6], where hailpad data was analyzed for the period 2008–2019, revealing that in 48% of the hail days, one to two hailpads were hit, while 15.2% of the days were associated with more than 10 hailpads being hit. June was the month when hail days were the most frequent, although the maximum number of hit hailpads occurred in April, with a second maximum in June. This is because in April 2023, the highest monthly number of hailpads in the studied period were hit (70 hit hailpads), and this number contributed to the high sum of the month. Results for April in the study period merit special consideration. A large fluctuation in the number of hailpads was recorded during the month of April, with high values observed every other year (2019 with 40 hailpads, 2021 with 55, and 2023 with 70 hailpads), and very low values recorded in the intervening years (below 10 hailpads or even zero). Years with high values of hit hailpads are characterized by a mild preceding winter, which led to relatively warm surface layers in April. As a result, the environment favored convective activity early in the season, which was released easily when the appropriate upper atmosphere weather systems were approaching. This finding for recent Aprils contradicts the previous statistical data of the GNHSP, which revealed April to be the month with the second-lowest occurrence of hail after September [3,6]. This outcome could be regarded as a hint of the influence of climate change.

A monthly analysis was conducted, examining the dataset according to the three size categories. Pea-size hail days predominated in the transition months (April, May, and September) and in June, while grape-size hail days predominated in the pure summer months (July and August) (Figure 2a). Finally, the occurrence of walnut-size hail is clearly the least frequent, being almost evenly distributed throughout the years and presenting a well-pronounced maximum in June, a month which traditionally combines, to the maximum degree, both dynamic and thermal forcings. The results for hit hailpads per size and per month also reveal interesting features. Pea-size hailpads are the most frequently encountered in all months, with their maximum value recorded in April (Figure 2b). Hailpads with grape–walnut size hailstones are found with higher frequency in June, followed by August and July. These findings show that hailfalls during the spring months infrequently contain very large hailstones, and that a prerequisite for their occurrence seems to be the presence of significant thermal heating.

An analysis was also performed regarding the distributions in space for the study period of (a) the frequency with which each hailpad was hit; (b) the sizes of hailstones recorded on each hailpad; and (c) the cumulative kinetic energy per hailpad, which directly combines both the previously mentioned parameters, presented on the maps of Figure 3 and Figure 4. In these maps, each dot corresponds to a hailpad, but only hailpads that have been hit are shown. The dots increase in size as the depicted parameters’ values increase following the increase in the classes they are classified in.

The frequency with which each hailpad out of the 157 in the network was hit was divided into five classes within the six-year study period: 1–2, 3–4, 5–6, 7–8, and 10–12 times (Figure 3a). Every hailpad of the network was hit at least one time. The maximum frequency was 12, recorded for only one hailpad, followed by a frequency of 11 for another hailpad. These hailpads are located at the northwest part of the protected area. Clearly, the region most frequently affected by hail is the northwest-west part of the protected area, and locally, the southern part. Figure 3b shows the hailstone size distribution. Pea-size hail (black dot) is distributed almost evenly over the whole area. Grape-size hail (red triangle) is also almost evenly distributed, apart from the eastern part of the area adjacent to the seacoast, where hail of this size is less frequently recorded. Walnut-size hail (green circle) is mainly recorded over a rather narrow strip crossing the entire area from northwest to southeast in its central part.

Finally, the cumulative kinetic energy on every hailpad for the whole study period was categorized into five classes, and its spatial distribution is presented in Figure 3c. In general, the northern part, as well as the southernmost part, receive high amounts of cumulative kinetic energy, and they are generally the same regions that are most frequently affected by large-size hail, as expected.

For a more detailed insight, the above spatial distributions were examined per month. The goal was to pinpoint any possible differences in hailfalls throughout the hail suppression season, and cumulative kinetic energy was selected as the most indicative parameter of destructive hail behavior. Monthly cumulative kinetic energy was also classified into five classes between 0.03 and 62.48 J as in Figure 4, with the total of them recorded only in June, providing clear evidence that the most intense hailfalls occur during that month. In September, on the other hand, only the first two lower classes are shown (0.03–3.66 and 3.66–8.63 J), as hailfalls during that month are rarer and of smaller size. In May, only the first three classes of the cumulative kinetic energy are present (0.03–3.66, 3.66–8.63 and 8.63–17.45 J). Finally, the remaining months of the hail season (April, July, and August) present similar results, with the first four classes of cumulative kinetic energy present, but for different reasons. In April, this is due to the high frequency of the hailfalls, especially of small size, while in July and August, this is due to larger-size hailfalls, although rarely observed.

The spatial distribution of cumulative kinetic energy per month also reveals significant variations. For each month, we identified demarcated areas of hail or no hail occurrence. The greatest hail coverage was observed in April. In May, the corresponding area is mainly limited to the central-western part, while in June, it is more extensive, with the higher values in the northern part. The maximum of the cumulative kinetic energy distribution clearly shifts eastwards in July, while in August, it reappears in the north. In September, the hail-covered area is the most limited of all the months, and it lies in the south. The small, most southeastern part of the project area presents unique results, with infrequent but intense hailfalls, observed only in April, July, and August.

The monthly cumulative kinetic energy distribution could be related quite satisfactorily with the position of the jet stream and its movement relative to the area depending on the season of the year. Its position is mainly located south of the area in the spring months, while in the summer months, it is mainly in the north. In cases of dynamic forcing during the summer, such as cases of longwave troughs and closed lows, the location of the jet stream varies, and it is temporarily located south of the area, or west, introducing a SW flow and producing favorable conditions for weather activity. In the summer months of July and August, the occurrence of hail is more local, but intense, compared to the spring months of April and May, where it is more widespread but of smaller size. Another significant factor which seems to be related to the larger-size hailfalls during the summer months is the convergence of the warm, moist air in the form of an easterly sea breeze coming from the Thermaikos Gulf with the descending air from the western mountain range in the center of the area, leading to an acceleration of air currents.

4. Conclusions

A hailpad dataset is used for the purpose of the study to define the temporal and spatial distributions of selected hail parameters in Central Macedonia.

June is the month when hail days are most frequent, although the maximum number of hit hailpads occurred first in April, and then in June. Pea-size hail days predominate in the transition months (April, May and September), while grape-size hail days predominate in the pure summer months (July and August). Pea-size hailpads are most frequently encountered in all months, with their maximum value recorded in April. Hailpads with the largest hailstones are found with greater frequency in June, followed by August and July.

Clearly, the region most frequently affected by hail is the northwest-west part of the area, and locally, the southern part. Pea- and grape-size hail is distributed almost evenly over the whole protection area, but walnut-size hail is mainly recorded over a rather narrow strip that crosses the entire area from the northwest to southeast in its central part.

Intramonthly, the greatest hail coverage is observed in April, and the lowest in September. In May, the corresponding area is limited to the central-western part, while in June, it is more extensive, with higher values in the northern part. The hail preference area clearly shifts eastwards in July, while in August, it reappears in the north.

These findings show that hailfalls during the spring months infrequently contain very large hailstones, and that a prerequisite for their creation seems to be the existence of significant thermal heating. The occurrence of walnut-size hail is clearly the least frequent, with a maximum in June, a month when both dynamic and thermal forcings are maximally combined. The spatial hail distribution could be related quite satisfactorily with the jet stream position and its movement relative to the project area, depending on the month of the year. In general, the geographical characteristics of the area are favorable for hail occurrence. It is a plain surrounded by high mountain ranges to the north and west, and constitutes lowland to the east, which facilitates the influence of the sea breeze with the advection of moist air masses. Convergence is the trigger for the release of instability in the afternoon hours due to heating. When a dynamic cause is also present, the occurrence of storm activity intensifies, producing extended hailfalls and, locally, large hailstones.