Research into Meteorological Drought in Poland during the Growing Season from 1951 to 2020 Using the Standardized Precipitation Index

Abstract

Meteorological drought (MDr) causes considerable economic losses in many countries, including in relation to agriculture. To examine the diversity of seasonal meteorological drought in Poland, the study uses monthly precipitation sums in the period from March to November, collected from 74 ground-based meteorological stations in 1951–2020. The paper defines meteorological drought on the basis of the standardized precipitation index in three seasons (SPI-3) and differentiates three degrees of drought intensity: extreme, severe and moderate. The study also calculates the size of Poland’s area affected by meteorological drought (AAMDr) and determines the relationship between AAMDr and SPI-3. MDr in Poland occurred more frequently in spring and autumn (every 4–5 years) than in summer (every 7 years). In the areas affected by extreme drought, precipitation was below the average, mainly ≤50% of the climatic norm, and air temperature values were below or above the average of the climatic norm, mainly ranging from −1.0 to 1.0 °C. A significant negative correlation between AAMDr and SPI-3 indices was found. The obtained results could be useful for managing climatic risk and developing regional and local agriculture adaptation plans aimed at mitigating the effects of climate change.

1. Introduction

Meteorological drought (MDr) is the result of at least several climatic factors interacting with each other, including astronomical, geographical, circulational and anthropogenic [1,2,3,4,5]. In Poland, MDr occurs as a result of anticyclonic patterns of atmospheric circulation over Europe, which are blocked by low-pressure systems [6]. Cyclonic circulation types from the east and south-east are also conducive to the occurrence of drought, though to a much lesser extent [7]. In Poland, such weather conditions occur in March, May, June and October [8]. It was also found that there is a weak but significant link between ENSO and surface hydrology in SW Poland [9].

In the last decade, the frequency and intensity of meteorological droughts have increased in most parts of Europe, mainly in Southern and Western Europe, but also in the region of the Carpathians [10,11]. Northern Europe is an exception. Its conditions, as proved, are becoming more and more humid and thus the frequency and severity of drought have shown downward tendencies; Central and Eastern Europe have acted as a transition zone [10]. The research conducted on a world scale indicates upward trends for droughts in central and eastern parts of North America, Western Canada, South America, Europe, Central Asia, Russian Far East, South Asia, Southern Africa and Central West Australia [12]. The same research shows downward trends mainly for Southern Canada, the north of South America, Central and North Africa, parts of Europe (e.g., Germany and Turkey), the east of Australia and especially for eastern parts of China and Mongolia. On the other hand, McCabe and Wolock’s [13] research for 1901–2009 does not confirm a drought trend on a global scale; at the same time, it was proved that drought intensified in the northern high latitudes. In the Sixth Assessment Report of the IPCC [14], it was stated that with global warming of 2 °C and higher, several regions in Africa, South America and Europe would experience an increase in the frequency and/or intensity of agricultural and ecological droughts; growths are also forecast for Australasia, Central and North America and in the Caribbean. In Poland, which is included in the region of Central and Eastern Europe, it is predicted that meteorological drought will occur more frequently, particularly in spring [10,15].

Local, short-term meteorological drought usually causes temporary inconveniences that may be almost undetectable in ecosystems. Long-term MDr that affects entire regions of a given country has a greater impact on society and the economy [16]. The phenomenon of drought causes considerable damage and economic losses in the EU, which are estimated at more than 100 billion euros in the last 30 years [17,18]. These studies showed that between 1993−2013 as much as 80% of lower wheat yields in 25 member states of the EU, including Poland, were caused by drought and heat waves. In the EU, the mitigation policy of socioeconomic and meteorological drought effects was shaped by such documents as: the Water Framework Directive (WFD) 2000/60/EC, A Blueprint to Safeguard Europe’s Water Resources [19]. In the world, the effects of droughts in agriculture are mitigated through irrigation systems, provision of water resources by building dams, improvement of water retention capacity and also development of comprehensive interbasin water transfer. Short- and long-term drought mitigation methods in different regions of the world are a subject of research and practical guidelines [19,20].

MDr occurs unevenly throughout Poland: its intensity varies, and it is seasonal. Therefore, its impact should be considered in relation to the place and time of its occurrence, among other things [21,22,23]. In Poland, MDr may result in heavy agricultural losses, which are reflected in reduced harvests and poor-quality agricultural produce [24,25,26,27,28,29]. Negative effects of MDr increase in specific conditions of Poland’s geographical environment (uneven distribution of natural bodies of water: mostly in the north and a latitudinal arrangement of the main mountain ranges located in the south) and also as a result of wrong water management policies in the previous century (focused on land drainage for the purposes of agriculture and settlement of that time). In order to reduce water deficits, temporary limitations on the use of water for non-potable purposes are often introduced.

For this reason, monitoring of meteorological conditions is very important, particularly in such countries as Poland, in which agriculture is a significant sector of the economy, and where agricultural land accounts for approx. 60% of Poland’s area and forest land for another 30%. The monitoring of MDr uses data from ground-based stations, which are evenly distributed in particular regions of a country, as well as from satellite images [10,15,22,30]. In Poland, agricultural drought monitoring is conducted by the Institute of Soil Science and Plant Cultivation–State Research Institute in Puławy. During the growing season of plants, i.e., from 21 March to 30 September, it prepares 10-day reports about the values of the climatic water balance (CWB) for particular communes (3058 communes), taking into consideration water requirements of main plant species and plant groups cultivated in Poland for four categories of soils of different drought susceptibility [31]. Plant species and groups covered by drought monitoring of the institute include: winter and spring cereals, maize, oilseed rape, potato, sugar beet, hops, tobacco, field vegetables, fruit trees and bushes, strawberries and legumes. The information about agricultural drought is announced by the minister competent for agriculture from 21 May to 20 October. The legal basis for organizing drought monitoring is the provisions of the Act on the Crop and Livestock Insurance [32].

Meteorological drought can be determined based on meteorological indices, which help to evaluate its variations in time and space, its intensity and also sometimes its beginning and end [33,34,35,36]. One of the best (and simplest) indices that is used to identify MDr and to evaluate its intensity is the standardized precipitation index (SPI) [37,38]. SPI can be calculated in various time scales, e.g., SPI-1–a month, SPI-3–a season, SPI-6–half a year, SPI-12–one year and SPI-24–two years [39,40,41,42,43]. In 2010, the World Meteorological Organization recommended the SPI index for the monitoring of meteorological drought by national meteorological services [44]. Currently, SPI is used for drought monitoring by, for example, the USA National Drought Mitigation Center [41] and the European Drought Observatory [45].

The aim of this work is to find the temporal and spatial diversity of the value of the standardized precipitation index (SPI-3) during three Polish seasons corresponding to the growing season from 1951 to 2020: spring (March to May), summer (June to August) and autumn (September to November). The study also aims to determine how much of Poland experienced meteorological drought (AAMDr) in the analyzed multi-annual period. The paper also examines the relationship between the AAMDr and SPI-3 indicators.

2. Materials and Methods

The meteorological data are from ground-based monitoring of weather conditions operated by the Polish Institute of Meteorology and Water Management (IMGW-PB) (Figure 1).

The paper uses precipitation sums (Pr, mm) and average air temperature (Ta, °C) in three seasons: spring (from March to May), summer (from June to August) and autumn (from September to November). The Pr and Ta data were collected from 74 meteorological stations from 1951 to 2020. In rare cases of incomplete meteorological data (approx. 2% of the dataset), these were supplemented from IMGW-PIB stations that were located closest to and best reflected the geographical conditions of the area with incomplete data. To this end, this paper uses linear and non-linear regression equations that describe the relationships between the analyzed series of data in the researched seasons. The goodness of fit between the regression functions and the empirical data was assessed on the basis of the coefficient of determination (R², %) and the Student’s t-test. The homogeneity of seasonal Pr and Ta series was verified with Bartlett’s test at a significance level of α < 0.05 [46,47]. Average meteorological data values for Poland were calculated based on data from 74 IMGW-PIB stations. The reference period (climatic norm) was the multi-annual average from 1991 to 2020.

Meteorological drought in spring, summer and autumn was determined by means of the standardized precipitation index (SPI-3), calculated based on the following equation:

$$\mathit{SPI}\mathit{\text{-}}\mathit{3}=\frac{f\left(Pr\right)-\mu }{\sigma }$$

(1)

where:

• Pr–seasonal sum of precipitation;
• μ–average value of the normalized series of precipitation sums;
• σ–mean standard deviation of the normalized series of precipitation sums.

Prior to determining SPI-3, the series of seasonal precipitation sums was normalized with the use of a transformation function f(Pr) = ∛Pr [32].

SPI-3 reflects short- and medium-term humidity in an environment. This indicator enables seasonal estimation and classification of precipitation variation, and it is frequently applied for the purposes of agriculture [44]. Meteorological drought classes are adopted from the classification proposed by [36,37], presented in

Table 1. Classes of meteorological drought according to the SPI-3 index.

Class	Symbol	SPI-3
Extreme drought	Ed	≤−2.0
Severe drought	Vd	[−1.5; −2.0)
Moderate drought	Md	[−1.0; −1.5)

.

The area of Poland affected by meteorological drought (AAMDr, %) was determined separately for each of the three seasons (spring, summer and autumn) according to the three SPI classes (Table 1) for each year from 1951 to 2020, i.e., for 210 seasons. AAMDr indicator calculations were performed in ArcGis 10.8.1. AAMDr was characterized using the following statistics: mean ($\overline{\mathrm{x}}$, %), standard deviation (sd, %), and extreme value (max, %). The highest value was adopted as the extreme value for the selected seasons and particular classes of meteorological drought in the analyzed multi-annual period of 1951−2020. The linear trend (r) at significance levels of α < 0.1 and α < 0.05 was determined with the use of the Pearson correlation coefficient.

Evaluation of the dependence of AAMDr on SPI-3 for Poland as a whole was performed by means of linear regression equations. For each of the examined seasons, six models were built for various ranges of the SPI-3 index. These ranges were determined in the range between the lowest and the highest value of the SPI-3 index in a given data series. The equations were verified by the coefficient of determination (R², %), the Student’s t-test, Snedecor’s F-distribution and the mean absolute error (MAE, %), which was determined with the formula:

$$MAE=\frac{1}{\mathrm{N}}{\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{N}}\left[\mathrm{yt}-\mathrm{Yt}\right]$$

(2)

where:

• yt–actual value;
• Yt–forecast value.

Using the inverse distance weighting method and ArcGis 10.8.1 software, maps were created that show the spatial distribution of SPI-3, Pr expressed as a percentage of the climatic norm (ΔPr, %), Ta expressed as a deviation from the climatic norm (ΔTa, °C), and the frequency of meteorological drought. This method was also used by [30]. The spatial resolution of the maps of the administrative division of Poland (NUTS 2) was 4 × 4 km. Statistical analyses were performed using STATISTICA 13.3. software and Excel 2010.

3. Results

3.1. SPI-3 Index–Spatial and Temporal Distribution

Out of the three analyzed seasons, extreme drought (SPI-3 ≤−2.0) occurred most frequently for the whole country in autumn in five years (1951, 1959, 1982, 2005 and 2011); this is followed by spring in three years (1953, 1959 and 1974); and least often in summer, when it occurred in two years (1992 and 2015) (Figure 2).

The lowest SPI-3 values were calculated for spring in 1953 (SPI-3 = −2.2), for summer in 2015 (SPI-3 = −2.4) and for autumn in 1959 (SPI-3 = −3.2). Severe drought in Poland, SPI-3 ϵ [−1.5; −2.0), was recorded with the following frequency: in five years in spring (1957, 1964, 1993, 2011 and 2018), in four years in summer (1976, 1983, 1994 and 2019) and in one year in autumn (1953). Moderate drought, SPI-3 ϵ [−1.0; −1.5), was recorded 8 times in transitional seasons in spring (1956, 1963, 1976, 1982, 1988, 2003, 2012 and 2020) and autumn (1954, 1961, 1969, 1985, 1986, 2000, 2014 and 2018), but only four times in summer (1951, 1952, 1982 and 2003). In transitional seasons, MDr occurred every 4–5 years for the whole country and every 7 years in summer.

Meteorological drought in Poland occurred in all the examined seasons in only one year: 1982. In this year, the drought was moderate in spring (SPI = −1.2) and summer (SPI = −1.5), and it was extreme in autumn (SPI = −2.7) (Figure 2). In seven of the 70 analyzed years, MDr of varying intensity was recorded in two seasons: in spring and summer in 1976 and 2003; in summer and autumn in 1951; and in spring and autumn in 1953, 1959, 2011 and 2018.

According to SPI-3 calculated for the whole of Poland, extreme drought occurred in three spring seasons. However, its intensity and occurrence across particular regions of the country in these seasons was very diverse (Figure 3). In the spring of 1953, extreme drought was recorded in east Poland; in 1959, extreme drought was recorded in smaller areas located in the northwest, central and southeast parts of the country; in 1974, extreme drought was recorded mainly in the north of Poland. In the areas with SPI-3 ≤−2.0, the recorded precipitation sum was lower than the average, mainly ≤50% of the climatic norm (ΔPr). At the same time, air temperature reached values below or above the average of the climatic norm (ΔTa) and generally fluctuated within the following ranges: from −2.0 to −0.5 °C in 1953, from −0.5 to 0.5 °C in 1959 and from −1.5 to −1.0 °C in 1974.

In summer, MDr of extreme intensity occurred in 1992 in northeast Poland and in 2015 in east and south Poland (Figure 4). Meteorological drought was recorded at ΔPr ≤ 50% of the climatic norm and ΔTa ≥ of the climatic norm, fluctuating between 0.0 and 1.0 °C.

In autumn, like in the other two analyzed seasons, extreme meteorological drought occurred in different parts of Poland (Figure 5). In 1951, extreme drought was identified in north and central Poland; in 1959, MDr was identified in almost the whole of Poland, except for the northeast part; in 1982, it was recorded in west Poland. In 2005, extreme drought was recorded only in small areas in the central and southern parts of the country; in 2011, extreme drought was identified in east, central and southeast Poland. During extreme drought, ΔPr was most often within 0–50% of the climatic norm; however, ΔTa oscillated between <−2 °C in 1959 and >1.5 °C in 1982.

The monitoring of MDr and consequently the assessment of its impact on agriculture takes into account not only the intensity but also the frequency of this phenomenon. The frequency of meteorological drought is not uniform across Poland and depends on the season (Figure 6). In spring, MDr identified on the basis of SPI-3 occurred most frequently in north (from 30% to 35%) and east Poland (from 25% to 30%); in summer, drought was identified mainly in north and central-west Poland (from 15% to 20%); in autumn, drought occurred primarily in northwest and south Poland (from 20% to 30%).

In the spring season, SPI-3 displays a significant (at a level of α < 0.1) negative trend in 1951–2020, which means that year-by-year MDr will intensify, mainly in southwest Poland (Figure 7).

In summer, a negative SPI-3 trend is visible in the majority of the country, including the entire southern and central Poland. In the autumn season, a negative SPI-3 trend is confirmed only in single locations, mainly in the central part of the country. A significant (at a level of α < 0.1) positive SPI-3 trend was found in the spring season in east Poland, and also in autumn, but only in single places in the south and north of Poland.

3.2. AAMDr Index–Temporal Distribution

The area of the country affected by meteorological drought (AAMDr), calculated on the basis of SPI-3, varied and fluctuated on average from 11.1% in summer to 18.9% in spring (

Table 2. Characteristics of Poland’s area affected by meteorological drought (AAMDr), determined on the basis of SPI-3. Years 1951–2020.

Season	Intensification of Meteorological Drought	Characteristics
		$\overline{\mathrm{x}}$ (%)	sd (%)	Max (%)	r for a Linear Trend
Spring	ed	2.1	6.0	34.6	–0.297 **
	vd	4.7	7.7	38.7	–0.247 **
	md	12.1	13.7	57.0	–0.067
	MDr (ed + vd + md)	18.9	23.3	83.8	–0.198
Summer	ed	1.1	4.0	26.4	0.137
	vd	3.4	8.9	42.1	0.115
	md	6.6	10.3	37.7	0.024
	MDr (ed + vd + md)	11.1	21.2	93.4	0.085
Autumn	ed	3.9	13.9	76.7	–0.111
	vd	4.2	9.7	50.1	–0.088
	md	7.7	11.6	57.0	–0.182
	MDr (ed + vd + md)	15.8	27.6	99.5	–0.228 *

Notes: ed–extreme drought, vd–severe drought, md–moderate drought, MDr–meteorological drought, $\overline{\mathrm{x}}$–mean (%), sd–standard deviation (%), max–the highest value in the multi-year period (%), r–Pearson correlation coefficient, ** significant at α ≤ 0.05, * significant at α ≤ 0.1.

).

Extreme drought affected on average only 2.4% of Poland’s area: in autumn, extreme drought affected the largest area, 3.9%; in spring, it affected 2.1%; and in summer it affected 1.1% of Poland’s area. The area affected by severe drought was on average 1.7% bigger than the area affected by extreme drought. The area affected by severe drought in 1951–2020 oscillated between 3.4% in summer and 4.7% of Poland’s area in the spring season. Moderate drought affected, on average, the largest part of Poland’s area in the spring season (12.1%). In spring, the area affected by drought was approx. 1.5–2 times bigger than in summer (6.6%) and autumn (7.7%). The biggest standard deviation was found for AAMDr of moderate intensity in both spring and summer, when it amounted to 13.7% and 10.3%, respectively. The situation looked different in the case of the autumn season because extreme drought was the most variable (sd = 13.9%), followed by moderate drought (sd = 11.6%). For AAMDr of extreme and severe intensity, in the spring season a significant (at a level of α < 0.05) linear trend was found; the Pearson correlation coefficients were −0.297 for extreme drought and −0.247 for severe drought. It can be seen that the area of Poland affected by meteorological drought also decreases year by year in autumn (r = −0.228, α < 0.1).

The percentage area of Poland that is affected by meteorological drought in spring fluctuated from 0% in 1962, 1967, 1994, 1995, 1997, 1998, 2009, 2010 and 2014 to as much as 83.8% in 1959 (Figure 8).

The highest AAMDr values with the extreme intensity of droughts were recorded in 1953 and 1974, when they amounted to 34.6% and 31.1% of the area of Poland, respectively. AAMDr of severe intensity amounted to 38.7% in 1959 and 24.4% in 1974. The highest values of AAMDr, >40% country’s area, were related to moderate drought. AAMDr values of >40% occurred in four years: 1993 (57.0%), 1976 (47.6%), 1957 (45.9%) and 2018 (40.1%).

The value of AAMDr in summer fluctuated from 0% in 1956, 1960, 1965, 1974, 1980, 1981, 2000, 2001, 2006, 2010, 2011 and 2016 to as much as 93.4% in 1992 (Figure 8). Apart from 1992, AAMDr was also >70% in another three years: 2015 (84.1%), 1994 (73.8%) and 1983 (73.0%). The biggest area of Poland affected by extreme drought was found for the following years: 2015 (26.4%), 1983 (14.7%) and 1992 (13.5%). Severe drought affected the largest area in 1992 (42.1%), 1994 (36.0%), 1983 (33.2%) and 2015 (32.9%), whereas the area with moderate drought was largest in 1992 (37.7%), 1976 (37.4%) and 2019 (36.4%).

The largest area of the country, >90%, affected by meteorological drought in autumn was recorded in 5 years: 1959 (99.5%), 1951 (98.9%), 1982 (96.8%), 2005 (92.4%) and 2011 (91.0%) (Figure 8). The AAMDr index of extreme intensity was highest in 1959 (76.7%), followed by 2011 (60.8%), 1951 (48.7%) and 1982 (46.5%). In turn, the largest area of Poland affected by severe drought was recorded in 2005 (50.1%) and 1969 (36.6%). The largest area of Poland affected by moderate drought was recorded in 1954 (57.0%), 1953 (40.0%) and 2005 (37.2%).

3.3. Dependence of AAMDr on SPI-3

Examination of the dependence of the country’s area affected by meteorological drought (AAMDr) on SPI-3 showed a significant (α < 0.01) and close relationship between the analyzed indicators in each of the analyzed seasons (

Table 3. Dependence on SPI-3 of Poland’s area affected by meteorological drought (AAMDr) in its selected ranges in spring, summer and autumn. Years 1951–2020.

Spring
SPI-3	N	R2	F	t	MAE	α
–2.2 ÷ 2.4	70	0.686	149.1	–12.2	10.4	0.01
–2.2 ÷ 1.5	66	0.773	218.0	–14.8	9.6	0.01
–2.2 ÷ 1.0	64	0.800	248.7	–15.7	9.0	0.01
–2.2 ÷ 0.5	55	0.852	304.8	–17.5	7.8	0.01
–2.2 ÷ –0.0	43	0.875	287.2	–16.9	6.6	0.01
–2.2 ÷ –0.5	32	0.849	169.2	–13.0	7.1	0.01
Summer
SPI-3	N	R2	F	t	MAE	α
–2.4 ÷ 1.9	70	0.645	123.8	–11.1	10.1	0.01
–2.4 ÷ 1.5	68	0.682	141.4	–11.9	9.8	0.01
–2.4 ÷ 1.0	59	0.769	190.5	–13.8	8.6	0.01
–2.4 ÷ 0.5	52	0.814	219.4	–14.8	8.1	0.01
–2.4 ÷ –0.0	30	0.921	329.6	–18.2	5.5	0.01
–2.4 ÷ –0.5	15	0.902	119.0	–10.9	7.6	0.01
Autumn
SPI-3	N	R2	F	t	MAE	α
–3.2 ÷ 2.5	70	0.666	136.0	–11.7	12.5	0.01
–3.2 ÷ 1.5	67	0.758	203.8	–14.3	11.7	0.01
–3.2 ÷ 1.0	64	0.781	221.6	–14.9	11.4	0.01
–3.2 ÷ 0.5	50	0.858	290.4	–17.0	9.7	0.01
–3.2 ÷ –0.0	39	0.899	332.7	–18.2	8.2	0.01
–3.2 ÷ –0.5	27	0.886	191.4	–13.9	9.2	0.01

Notes: N–degrees of freedom, R²–coefficient of determination (%), F–Snedocor test, t–Student’s t-test, α–significance level, MAE–mean absolute error (%).

, Figure 9).

The SPI-3 index explained the variability of AAMDr in spring, from approx. 69%, when SPI-3 was within the range from –2.2 to 2.4, to as much as approx. 88%, when SPI-3 was within the range of from –2.2 to 0. The Student’s t-test and Snedecor’s F-distribution values that were calculated for the linear regression equations ranged respectively from –12.2 to –17.5 and from 149.1 to 304.8.

In summer, as in spring, the closest relationship between AAMDr and SPI-3 was found when SPI-3 ranged from −2.4 to 0. The coefficient of determination calculated for this equation amounted to as much as 92.1%; the value of the Student’s t-test was −18.2, and the value of Snedecor’s F-distribution was 329.6. In autumn, the dependence of AAMDr on SPI-3 was negative, significant at a level of α < 0.01, as in the case of spring and summer. R² ranged from approx. 67% to 90%, whereas the value of the Student’s t-test fluctuated from −11.7 to −18.2 and the value of Snedecor’s F-distribution ranged from 136 to 332.7. MAE ranged from 6.6% to 10.4% for spring, from 5.5% to 10.1% for summer and from 8.2% to 12.5% for autumn. The highest negative regression coefficient for the dependence of AAMDr on SPI-3, when SPI-3 ranged from the minimum value to 0, was found for the summer season (−39.62) and was lowest for the autumn season (−38.25).

4. Discussion

Meteorological drought is an important atmospheric phenomenon which limits agricultural development in many countries, causing, e.g., reduced yields of cultivated plants [33,48,49,50,51]. Therefore, MDr remains relevant, despite the fact that it has been the subject of many research studies [35,42,52]. However, it is still difficult to predict meteorological drought even in countries with very good weather monitoring [53,54]. Examination of meteorological drought is usually conducted after the event on the basis of historical materials [23,55,56].

In this paper, MDr was identified by means of the standardized precipitation index (SPI), whose advantages are its ability to determine drought intensity in three classes (extreme, severe and moderate) and the possibility of comparing results calculated for periods of different durations. This paper shows that in 1951–2020 the number of years with meteorological drought in Poland varied in particular decades and depended on the analyzed season. The research by [57] conducted for the years 1971–2015 confirms the occurrence of meteorological drought in Poland in the years in which it was also proved in this paper: 1974, 1992, 2005 and 2015. Ref. [56] showed the occurrence of drought in central Poland, on average, every 5 years, including extreme drought in only 2 years: 1982 and 1989. On the other hand, [21] identified extreme drought only in 1992. Ref. [50], with the use of the SPI-3 index, identified MDr in at least 3 of the 6 analyzed regions of Poland in the April–June season only in 1976. Similar to the present paper, Ref. [58] identified MDr of extreme and severe intensity in the growing season in 1969, 1982, 1992, 1994, 1996, 1997, 2006, 2009 and 2015. The slightly different MDr frequency that was found in the current paper, in comparison with the above-mentioned works, is mainly the result of the following: examination of a different, usually shorter, multi-annual period; a different period, most often a month rather than a season; and different regions of Poland. By means of SPI, researchers identified fewer periods of the occurrence of meteorological drought in Poland than with the use of other indicators, such as Selyaninov’s hydrothermal coefficient or the relative precipitation index [56,58,59], which may be explained by the stricter criteria of the analyzed index.

Based on the trend of SPI-3, this research study shows that in the coming years in Poland increased droughts are likely, mainly in spring and summer. At the same time, locally in the east of Poland, drought intensity may decrease in spring. These results detail general drought trend directions determined for Europe [12,14,22,23].

The occurrence of meteorological drought in Poland has considerable spatial diversity; therefore, assessing drought on the basis of averaged values of indicators for the whole area of Poland, or even for a given region, seems to be incomplete and insufficient [58]. In Poland, drought occurrence is variable, and therefore, with the exception of mountain regions in the south of the country, it is not possible to definitively indicate areas which would not be exposed to the risk of drought. Ref. [60] found that extremely dry periods occurred more often in the western part of the investigated region, and moderately dry periods were more frequent in the eastern part of Poland. On the other hand, and similar to the current paper, Ref. [61] most frequently identified extreme drought in northwest and central-east Poland and moderate drought in central and southeast Poland. The spatial distribution of drought in Poland changes depending on the applied indicator and the adopted criterion of selecting periods with a given intensity of meteorological drought. Moreover, the different distribution of the frequency of drought in 1951–2020 that was found by other authors [58,60] is a result of the fact that this paper adopted a different reference period, i.e., 1990–2020.

In 1951–2020, on average, the largest area of the country affected by meteorological drought (AAMDr) was identified in spring (18.9%), followed by autumn. On the other hand, the smallest area of the country with meteorological drought was identified in summer; in spring the drought-affected area was approx. 8% bigger than in summer. The research by [23] showed that, in the spring season, precipitation deficits affected, on average, 33% of Poland’s area, i.e., 14.1% more than shown in this study. In Poland, there is a shift in the occurrence of meteorological drought from summer to spring and autumn [47]. Most probably, this is an effect of clearly accelerated climate warming in the winter–spring season of the last 35 years [62] and the progressive disappearance of snow cover in winter, which results in declining water levels in soil in spring [63,64]. The occurrence of meteorological drought is a result of increasingly common short-lasting storms, often separated by long, dry periods, thus leading to atmospheric drought [65]. The research on meteorological drought contributes to improved drought monitoring, which fits in with implementation of some of the aims of the common agricultural policy in the European Union, i.e., supporting sustainable management of natural resources and fighting against climate change [66].

5. Conclusions

The paper shows that meteorological drought (MDr) in Poland occurs more frequently in spring and autumn (every 4–5 years) than in summer (every 7 years). In 1951–2020, meteorological drought did not occur with the same frequency across Poland. In the spring season, MDr was most often recorded in the east and north (every 3–4 years); in summer, MDr most frequently occurred in north and central-west Poland (every 5–7 years); in autumn, MDr was most often recorded in the south and northwest of Poland (every 3–5 years).

Extreme drought (SPI-3 ≤−2.0) in spring across the whole of Poland occurred in 3 of the 70 examined years: 1953, 1959 and 1974. In summer, extreme drought was recorded in two years: 1992 and 2015. In autumn, extreme drought occurred in five years: 1951, 1959, 1982, 2005 and 2011. In the areas with SPI-3 ≤−2.0, precipitation was below the average, mainly ≤50% of the climatic norm (ΔPr), locally even ≤25%, which happened in the autumn season in three years: 1951, 1959 and 2011. On the other hand, air temperature values were below or above the climatic norm (ΔTa) and mainly ranged from −1.0 to 1.0 °C, whereas ≤−2.0 °C occurred only in 1959 in the autumn season.

In 1951–2020, the area of the country affected by meteorological drought (AAMDr), determined based on the SPI-3 index, amounted on average to 18.9% in spring, 11.1% in summer and 15.8% in autumn. AAMDr of extreme intensity ranged from 1.1% to 3.9%; the area affected by severe drought fluctuated between 3.4% and 4.7%; and the area affected by moderate drought ranged from 6.6% to 12.1%. The values of AAMDr ≥75% were calculated for the years when SPI-3 amounted to ≤−2.0. The spring season of 1993 was an exception, because as much as 57% of the country’s area was affected by moderate drought. In the autumn seasons of 1951 and 1959, almost all of the country’s area was affected by drought: in these years, respectively, extreme drought affected approx. 49% and 77% of Poland’s area; severe drought affected approx. 27% and 18%; and moderate drought affected approx. 24% and 5% of Poland’s area.

The paper proved a significant (at a level of α < 0.01) negative relationship between AAMDr and SPI-3; the closest relationship was found with SPI-3 within the range from the minimum value to 0. R² calculated for the regression equations ranged from approx. 88% to 92% and the values of Snedecor’s F-distribution ranged from approx. 287 to 333.

A positive SPI-3 trend was found in the areas of Poland with the highest frequency of meteorological drought in all of the examined seasons; this means that the risk of intensification of the analyzed phenomenon in these regions becomes reduced year by year. However, after considering scientific reports about a significant increase in air temperature in the last 30 years in Poland, it can be assumed that MDr will most probably intensify and expand.

The standardized precipitation index determined for the seasons makes it possible to evaluate short- and medium-term variations in meteorological drought in Poland in time and in space and also its intensity in three classes: extreme, severe and moderate. The research on the dynamics of meteorological drought change should be continued and even broadened with long periods of half a year, one year and two years, as the climatic instability that climatologists predict will occur by 2050 may significantly influence its development and intensity, which in turn will affect farming systems and agrarian structure in Poland.

The obtained results could be useful for managing climatic risk and developing regional and local agriculture adaptation plans that are aimed at mitigating climate change effects.