Characteristics and Motivations of Drought and Flood Variability in the Northern Haihe River Basin over the Past 500 Years

Abstract

The Haihe River system, located in the East Asian monsoon climate zone, experiences uneven precipitation and significant variability, leading to frequent droughts and floods that disrupted economic and social development. While many studies have assessed the risks of droughts and floods in the Haihe River Basin, most focus on the basin as a whole, leaving a notable gap in research on the dynamics of the northern region. This study analyzed historical drought and flood data, incorporating instrument precipitation records from 1960 to 2009 to reconstruct conditions in the northern Haihe River Basin from 1470 to 2009. Using methods like the Mann–Kendall test, sliding averages, continuous wavelet technology, and spatial analysis, this study examined the trends, change points, periodicity, and spatial patterns of drought and flood variability. The findings showed that from 1470 to 2009, drought and flood variabilities occurred 73.15% of the time in the northern Haihe system, with peak disaster periods in the 17th, 19th, and 20th centuries. The region has alternated between wet and dry cycles, with a notable dry trend emerging in the 21st century. A prominent 35~50-year cycle in drought and flood occurrences was identified, along with high-frequency oscillations. Flood periods were most frequent in the eastern plains, while drought periods were more prevalent in the western areas, gradually shifting eastward since 1950. The research also revealed correlations between drought and flood variability and solar activity, with peak years coinciding with higher frequencies of these events. El Niño events were associated with drought periods, while La Niña events tended to cause flood periods. Factors such as solar activity, El Niño–Southern Oscillation, monsoon climate patterns, topography, and human influences shaped the dynamics of drought and flood variability in the northern Haihe River Basin. A comparison with other regions showed consistent wet and dry cycles over the past 500 years, particularly between the northern and southern parts of the basin. However, since the 21st century, the southern region has remained humid, while the northern region has become increasingly drier. Despite similar temperature trends, humidity changes have diverged in the modern warming period. Although the underlying factors driving drought and flood variability were not fully understood and required a further exploration of the global climate system’s interactions, these findings emphasized the need for targeted strategies to address the ongoing challenges of drought and flood management in the northern Haihe River Basin.

1. Introduction

Drought and flood variability are among the most prevalent and destructive natural disasters worldwide [1,2,3,4,5,6]. In China, these events usually resulted in significant annual losses, including approximately 20 billion kilograms of food and over CNY 200 billion in economic damage [7,8]. As global climate change has accelerated, particularly rising temperatures, the frequency and severity of drought and flood variability have intensified, posing an escalating threat to sustainable economic development and public safety [5,9]. Understanding these shifts, their underlying causes, and developing effective mitigation strategies are essential for advancing climate change research and improving future water resource management and disaster forecasting [4,6].

The Haihe River Basin, spanning eight provinces, autonomous regions, and municipalities, covers a drainage area of 317,800 square kilometers and faces distinct challenges. Despite its favorable geography and economic assets, the region has struggled with high population density and a monsoon climate, leading to uneven precipitation distribution and frequent drought and flood variability, with drought being the more pressing concern. The frequency and severity of drought risk in this basin ranks among the highest in China [6,10,11,12]. Historically, the basin has been a cradle of Chinese civilization and a political and cultural center during the Yuan, Ming, and Qing dynasties, with extensive records documenting drought and flood events [6,13,14]. These records showed that, over the past 500 years, the Haihe River has experienced 135 years of widespread flooding, including 84 consecutive years of flooding lasting 2–4 years. In contrast, this region has faced 140 years of drought, with 90 consecutive years of drought during this period [14]. For instance, Xiao analyzed drought and flood patterns from 1470 to 1985, identifying periodic and persistent trends, with a notable increase in drought occurrences over the past century [15]. Sha and Guo studied the spatiotemporal distribution of major droughts and floods in the Haihe River Basin over the past centuries, identifying 26 basin-wide droughts and 29 major floods in 525 years, with cycles of approximately 2–3 years, 5 years, 10 years, and 20 years [16]. Using t-tests, the power spectrum method, and wavelet analysis, Lu et al. examined drought and flood cycles in the Haihe River Basin based on 540 years of data, revealing multiple fluctuation cycles, with a dominant 35-year cycle [17]. Wang et al. analyzed climate change trends in the Haihe River Basin over the past 51 years using Mann–Kendall nonparametric tests, sliding T-tests, and wavelet analysis, observing a decreasing trend in precipitation with cyclical changes in 5–7-year and 10–15-year intervals [18]. Wang and Hu used a regional drought and flood disaster index model and the empirical mode decomposition (EMD) method to investigate drought and flood patterns over the past 500 years, showing drought–flood–drought fluctuations with interannual cycles of 2.5, 9.8, 31.3, and 83.5 years, along with large-scale cycles of around 100 years [19]. Li et al. employed the Variable Infiltration Capacity (VIC) model to investigate the link between precipitation changes in the Haihe River Basin and the East Asian monsoon., revealing a similar 40-year cycle [20]. Since the 1970s, however, the weakening of the East Asian monsoon has led to prolonged drought conditions and water shortages in the Haihe River Basin.

Despite these efforts to understand drought and flood cycles and their historical context [15,16,17,18,19,20], there remains limited research focused on the specific conditions within the northern Haihe River system. Existing studies have primarily concentrated on hydrological assessments, precipitation trends, and drought–flood patterns over the past 50 years [11,21,22], often relying on historical records that may contain gaps or inaccuracies due to data limitations [23]. Furthermore, many analyses focused on broad temporal scales, complicating the understanding of the underlying mechanisms driving these disasters. While numerous studies have examined the risks of droughts and floods in the Haihe River Basin, most treat the basin as a whole [10,18,23,24,25,26,27], leaving a significant research gap concerning the dynamics of the northern regions.

This article aims to address this gap by focusing on the northern Haihe River system, offering a comprehensive analysis of historical drought and flood variability in the region from 1470 to 2009. By employing well-established research methodologies from both domestic and international studies, this investigation facilitates comparisons with drought and flood patterns in other regions. The findings of this article will be crucial for improving water resource management, fostering sustainable development, and implementing effective strategies to mitigate the impacts of drought and flood risks in the area.

2. Study Area

The northern section of the Haihe River system spans a vast area between longitudes 111° and 119° E and latitudes 38° and 42° N. Key cities within this region include Chengde, Zhangjiakou, Langfang, and Tangshan in Hebei; Datong, Shuozhou, and Xinzhou in Shanxi; Ulanchabu in the Inner Mongolia Autonomous Region; and Beijing and Tianjin [21] (see Figure 1). This area encompasses a total of 64 districts and counties. Geographically, the northern Haihe River system is narrow and elongated in the east–west direction, stretching 556.6 km in length, while its north–south stretch covers a total area of 83,900 km². The mountainous terrain occupies approximately 52,700 km², while the plains span around 31,200 km², yielding a mountainous-to-plain area ratio of about 1.69:1. The northern section of the Haihe River system serves as a secondary water resource zone within the Haihe River Basin, as illustrated in Figure 1a.

Situated within the East Asian continental monsoon climate zone, this region experiences a warm-temperate humid or semi-humid climate [28]. It is characterized by some of the lowest precipitation levels along China’s eastern coast. During winter, Siberian air masses dominate, bringing cold, dry conditions with minimal rain and snow. In spring, warmer temperatures, influenced by Mongolian air masses, combine with high winds, leading to significant evaporation and increased agricultural water demands, often resulting in early spring droughts. Summer is marked by heat and concentrated rainfall, while autumn sees cooler temperatures and reduced precipitation as the Mongolian high-pressure system strengthens. Precipitation and runoff are highly uneven, with 75% to 85% of annual rainfall and 50% to 90% of surface runoff occurring during the flood season (June to September). Interannual precipitation variability can exceed 30% in cities like Beijing and Tianjin. The combination of mountainous terrain, limited flood discharge capacity, and variable river flow contributes to frequent droughts and floods, significantly impacting regional economic sustainability.

The northern Haihe River system is composed of four major sub-basins: Yongding River, Jiyun River, Chaobai River, and Beiyun River, as shown in Figure 1a [23]. Among them, the Yongding River has the largest drainage area, formed by the confluence of the Sanggan River to the south and the Yanghe River to the north. Due to the significant elevation difference in the upper reaches, along with low vegetation coverage, strong erosion, and severe soil and water loss, the river appears yellow. The Jiyun River, Chaobai River, and Beiyun River flow roughly in parallel near their outlet to the sea and were regulated by gates and dams in the 1960s, connecting them into a unified system which is now known as the North Three Rivers.

Topographically, the Haihe River Basin is higher in the northwest and lower in the southeast, featuring a mix of plateau, mountain, and plain landforms [23,28] (Figure 1b). The Shanxi Plateau and Taihang Mountains dominate the west, with altitudes exceeding 1000 m, while the eastern plains, shaped by river alluvial deposits, are generally below 50 m in elevation (Figure 1b). The Haihe River ultimately flows into the Bohai Sea through these plains. Geologically, the study area spans from the Archean to the Quaternary period, encompassing a variety of rock types, including metamorphic rocks, limestone, and clastic sediments. This region is part of the North China Fault Block, characterized by alternating NNE-trending anticlines and synclines. Tectonic activity has shaped sediment accumulation patterns, with Quaternary deposits primarily consisting of loess and river sediments.

3. Materials and Methods

3.1. Materials

The study of historical climate change in China, particularly within the Haihe River Basin, provides a comprehensive framework for understanding droughts and floods over the past 5000 years. Chu’s classification into four periods, including archeological, phenological, local chronicles and instrument observation, highlights the evolution of climate records, transitioning from sparse archeological evidence to detailed historical accounts [29]. In the northern Haihe River region, significant records of climate events primarily fall within the phenological and local chronicles periods (1100 BC to 1900 AD), a time when human activities intensified, leading to a rich accumulation of cultural and historical documentation. The earliest records of droughts and floods date back to 781 BC, although early documentation was limited and uneven across different areas [28]. From 960 AD onward, the ‘Historical Materials on Natural Disasters in the Haihe River Basin’ became essential for assessing the impacts of disasters [23,28]. While data from 960 to 1469 AD were sparse, local chronicles from 1470 onward provided more comprehensive and reliable information regarding drought and flood periods [23].

The research is divided into two main periods. The first period, from 1470 to 2000, focuses on compiling the literature related to meteorological disasters from various local annals, which provide relatively complete records [14,23,28,30,31]. This phase primarily gathers and documents the literature on meteorological disasters, sourcing records from county annals, government archives, local histories, general chronicles, and water conservancy records, all offering comprehensive accounts. For example, the yearly charts of dryness/wetness in China over the past 500 years provide annual data on drought levels [14]. Additionally, yearbooks and county annals help validate the accuracy of historical drought conditions [23,24,25,26,27,28]. The second period, spanning from 2000 to 2009, focuses on annual precipitation data during the flood season, utilizing information from the China Meteorological Data Network. These data are used to analyze the variability of drought and flood severity through calculated precipitation indices [6,22,23,26]. This structured approach deepens the understanding of historical climate events and their impacts, contributing to the development of more accurate climate models and enhanced disaster preparedness strategies.

3.2. Drought-Level Identification and Rectification

Based on historical periods, current administrative divisions, and prior research on drought and flood variability in the Haihe River Basin, this article identified seven key cities in the northern Haihe River system as primary assessment stations for average drought and flood levels: Beijing, Chengde, Tangshan, Zhangjiakou, Tianjin, Datong, and Xinzhou (see Figure 1). Using the historical literature and contemporary precipitation data, this article classified annual drought and flood levels at these stations from 1470 to 2009, establishing a foundation for a comprehensive analysis of drought and flood patterns in the basin (referring to [6,14,23]). Given the multi-provincial nature of the research area, the timing of drought and flood events varied across the stations, complicating efforts to calculate an average for the entire watershed. This study employed a weighted average method, using the cultivated land area of each zone as a weighting factor, to establish the overall drought and flood levels for the watershed.

To effectively assess these levels in the Hebei region, confirming the suitability of the selected sites is crucial. This study compiled monthly precipitation data from the seven stations during the flood season (May to September) from 1960 to 2009, calculating the average precipitation over the past 50 years for the basin. Using mathematical analysis software R (version 4.2.2), the correlation coefficient between the flood season precipitation at each station and the basin’s average precipitation was calculated, ranging from 0.65 to 0.87. High correlation coefficients indicated a strong positive relationship between flood season precipitation at each station and the overall basin average. Building on these data, a multiple linear regression analysis was performed on the average dry and wet variations for each station and region within the study area. Finally, the resulting equation was used to reconstruct the comprehensive drought and flood level series for the basin. This average drought and flood level not only reflected the severity of disasters in a given year but also provided insights into the scale of the affected area.

In the ‘yearly charts of dryness/wetness in China for the last 500-year period’, drought and flood situations were initially categorized into five levels [14]. Subsequent studies have expanded this classification to seven or even nine levels. Acknowledging the low resolution of a five-level classification, this study derived a seven-level standard for the basin average, following the guidelines established by the Hebei Province Drought and Flood Forecasting Research Group [28]. Consequently, drought and flood levels in the Haihe River Basin were classified into seven categories, ranging from severe flood (level 1) to extreme drought (level 7) [23]. Due to limitations in mathematical processing methods, discrepancies might arise between calculated results and actual conditions, which could hinder accurate reflections of drought and flood disasters. Therefore, it was essential to verify these levels and make necessary adjustments. Special attention was given to years classified at levels 1, 2, 6, and 7, with the authenticity of these events assessed through the historical literature review, allowing for corrections of any identified discrepancies [23,28].

3.3. Methods

Using the historical literature and meteorological observation data, this paper reconstructed the drought and flood level series and examined the evolution of drought and flood patterns in the northern Haihe River Basin over the past 500 years [14,22,23,26,28,30,31]. This study employed a sliding average analysis, Mann–Kendall test, wavelet analysis, and spatial analysis to explore the correlation between drought and flood events, natural factors, and human activities. Additionally, it also compared drought and flood changes in other regions. Based on these findings, this paper proposed corresponding countermeasures for water resource mitigation strategies (Figure 2).

3.3.1. Sliding Average

The sliding average is a widely used method for smoothing fluctuations in variable changes, helping to identify underlying trends [23]. By selecting an appropriate sliding coefficient k, this method reduces the influence of short-term variations. For a sequence X with a sample size of n, the sliding average sequence is defined as

$$X_j={\displaystyle \frac{1}{k}}{\sum }_{k=1}^k{X}_{i+j-1}\dots (j=1,2,\dots ,n-k+1)$$

(1)

Here, k represents the sliding coefficient. A larger k emphasizes long-term trends over short-term fluctuations. In this study, k values of 5, 10, 20, and 50 are considered.

3.3.2. Mann–Kendall Test

The Mann–Kendall (M-K) test is a statistical method for assessing trends in time series data, including temperature, precipitation, and runoff [32]. Its strength lies in its reliance solely on the ranks of the observed values, making it robust to the actual data distribution.

For a time series X = {x₁, x₂, …, x_n}, the test statistic S is calculated as

$$S={\displaystyle {\mathsf{\sum }}_{k=1}^{n-1}}{\displaystyle {\mathsf{\sum }}_{j=k+1}^n}\mathit{sgn}(x_j-x_k)$$

(2)

where sgn(x) is the sign function.

$$sgn\left(x_j-x_k\right)=\left\{\begin{array}{c}1 x_j-x_k>0\\ 0 x_j-x_k=0\\ -1 x_j-x_k<0\end{array}\right.$$

(3)

The standard normal variable Z is computed as

$$Z=\left\{\begin{array}{c}{\displaystyle \frac{s-1}{\sqrt{var\left(S\right)}}} S>O\\ 0 S=0\\ {\displaystyle \frac{S+1}{\sqrt{var\left(S\right)}}} S<0\end{array}\right.$$

(4)

Here, n is the length of the sequence.

$$\mathit{var}(S)={\displaystyle \frac{1}{18}}n(n-1)(2n+5)$$

(5)

The null hypothesis H₀ can be rejected based on the calculated $Z_{\alpha /2}$ value. Critical values of 1.28, 1.64, and 2.32 correspond to significance levels of 90%, 95%, and 99%, respectively. Typically, a significance level of α = 0.05 is used; if a significant trend is detected at this confidence level, then |Z| > Z_α/2 = 1.64. A positive Z indicates an upward trend, while a negative Z signifies a downward trend.

For the time series X₁, X₂, …, X_n, we constructed a rank sequence $S_n\left(k\right)=V_k+V_{n-k}$ to represent the cumulative count of instances where X_i > X_j for 1 ≤ j ≤ i.

When the time elements form a randomly independent sequence, we can define the statistic similarly:

$$U{F}_k={\displaystyle \frac{\left(S_k-E\left[S_k\right]\right)}{\sqrt{\mathit{var}\left[S_k\right]}}}(k=1,2,3,\dots ,n)$$

(6)

For $x_t$(t = 1, 2, …, n), $V_k={{\sum }_{t=1}^k\left(x_t-{\overline{x}}_k\right)}^2$, $V_{n-k}={{\sum }_{t=k+1}^n\left(x_t-{\overline{x}}_{n-k}\right)}^2$, assuming a standard normal distribution, we evaluate the statistics for the time series X₁, X₂, …, X_n. The process can also be repeated in reverse chronological order. Given a significance level of α (e.g., α = 0.05), we plot the critical values alongside the curves of the two statistical sequences. If the inspection curve remains within the critical line, it suggests that the trend is not significant. A value greater than zero indicates an upward trend, while a value less than zero indicates a downward trend. If the statistic exceeds the critical line, it signifies a significant trend. If an intersection point occurs between the two curves at the critical line, this indicates when a significant change or mutation begins [23].

3.3.3. Continuous Wavelet Transform (CWT)

For a time series with n samples, the periodic function can be represented by its fundamental angular frequency [6,33]. The Fourier series in triangular form is expressed as

$$f\left(t\right)=a_0+{\sum }_{n-1}^{\infty }[a_n\cos \left(n{\omega }_{1}t\right)+b_n\mathit{sin}(n{\omega }_{1}t)]$$

(7)

where

$$a_0={\displaystyle \frac{1}{T}}{\int }_{t_0}^{t_0+T}f\left(t\right)d_t$$

(8)

$$a_n={\displaystyle \frac{2}{T}}{\int }_{t_0}^{t_0+T}f\left(t\right)\mathit{cos}(n{\omega }_{1}t)d_t$$

(9)

$$b_n={\displaystyle \frac{2}{T}}{\int }_{t_0}^{t_0+T}f\left(t\right)\mathit{sin}(n{\omega }_{1}t)d_t$$

(10)

$$D\left(f_n\right)=(a^2+b^2)n/2$$

(11)

After processing the data, we generated a spectrogram to visualize the frequency components, revealing multiple peaks that correspond to different periods of the time series. The frequency with the highest intensity indicates the dominant quasi-period of variation within the time series. This study employs MATLAB software (version R2024a) to conduct spectral analysis on drought and flood data, ensuring that the number of samples for the Fourier transform is a power of 2 for optimal results.

4. Results

The analysis of drought and flood data from 1470 to 2009 revealed a total of 395 disaster occurrences, indicating a frequency of 73.15%. This included 186 drought years and 209 flood years, yielding a drought-to-flood ratio of 1:1.12. Waterlogging disasters were recorded 38.70% of the time, averaging one occurrence every 2.58 years. Notably, severe floods (level 1) were recorded only once between 1949 and 2009, whereas severe drought periods became more frequent, especially from 1950 onward (Figure 3a).

The frequency distribution over the centuries showed variability, with significant drought and flood activities noted in the 17th, 19th, and 20th centuries. The 16th, 18th, and 19th centuries were generally wetter, whereas the 17th and 20th centuries saw more drought conditions (Figure 3b). The drought–wetness index (D/(D + F)) calculated for each century indicated a transition from wetter to drier periods over time (Figure 3b). Continuous drought and flood periods, meaning two or more consecutive years of such events, were documented over a span of 540 years, revealing 147 consecutive flood years and 142 consecutive drought years (Figure 3). Flood events persisted for up to nine consecutive years, while droughts lasted for seven consecutive years. The trends in annual drought and flood levels exhibited complex fluctuations, with significant oscillations, particularly in the mid-16th and early 20th centuries. Overall, while flood occurrences were slightly more frequent than droughts, increasing smoothing coefficients diminished this difference.

To facilitate a century-scale comparison and provide more detailed insights, the occurrence of drought and flood periods over the past 540 years was analyzed using a 50-year interval (Figure 3c). The findings showed that, in contrast to the century-long frequency trends, the overall 50-year changes were relatively gradual. Notably, the frequency of drought and flood variability varied significantly across these intervals. For instance, flood periods were 2–3 times more frequent than drought periods during the periods of 1551–1600 and 1751–1800. Conversely, drought variations were more prevalent than flood periods between 1601–1650 and 1951–2009.

The Mann–Kendall test revealed a slight increasing trend in drought and flood levels, suggesting a potential shift toward aridification, though this trend was not statistically significant (Figure 4a). A mutation analysis identified key turning points in 1533, 1620, and 1925, with a notable upward trend emerging around 1920 (Figure 4b). Drought and flood occurrences exhibited periodic characteristics, with a spectrum analysis revealing significant cycles of 12 years, 16 years, and 50 years (Figure 5). A wavelet analysis further confirmed the variations in oscillation periods over time (Figure 6). The wavelet analysis revealed a clear alternation between the positive and negative phases of drought and flood variability in the northern Haihe River system, particularly evident in the 16–30-year cycle. High-value centers were noted in 1665 and 1875, while a low-value center appeared in 1710. Additionally, multiple high- and low-value centers were seen around the 4–6-year cycle (Figure 5). Over the past 60 years, drought occurrences have notably increased, especially in Beijing and Langfang, underscoring a trend toward aridification in the northern Haihe River Basin. In contrast, flood frequencies have generally decreased, particularly from east to west, highlighting significant regional disparities.

Overall, the historical evolution of drought and flood periods in the northern Haihe River system showed a significant 35–50-year cycle. From 1470 to 1620, the cycle length was 5–8 years; from 1620 to 1740, it extended to 35–40 years; from 1740 to 1950, it was 12–16 years; and from 1950 to 2009, a high-frequency oscillation of approximately 5 years emerged. This indicated a trend toward shorter cycles of drought and flood variability over the past 50 years.

This article also provided a statistical analysis of drought and flood variability years and frequencies at seven stations in the northern Haihe River of China and included a frequency distribution map of these events (Figure 7). Figure 7a illustrated that flood frequencies in the northern Haihe River system were highest in the eastern plain areas, particularly in Tianjin and Tangshan, where the frequency reached approximately 40%. Over the past 540 years, these areas have experienced an average of 220 flood years, translating to an occurrence roughly every 2.25 years. Conversely, low-frequency zones were primarily located in the western areas, such as Xinzhou and Shuozhou, with frequencies below 20%. In these regions, flood years occurred an average of 132 times, or about once every 4.1 years. Figure 7b indicated that drought occurrences were concentrated around Zhangjiakou, exhibiting a frequency exceeding 40%. Over 540 years, drought events totaled 115, averaging once every 4.7 years. As one moved eastward from Zhangjiakou, the drought frequency decreased in a circular pattern, with the lowest being 23%. To the west of Zhangjiakou, drought frequencies remained high, with most areas (excluding regions near Datong) exceeding 35%, and a small high-frequency center emerging in Xinzhou.

5. Discussion

5.1. Natural Causes

5.1.1. Topographic Factors

The northern Haihe system, located between longitudes 111° and 119° E and latitudes 38° and 42° N, featured diverse topography with elevations rising in the west and descending toward the east. It was bordered by the Taihang Mountains and the Shanxi Plateau to the west, the Inner Mongolian Plateau and Yanshan Mountains to the north, the north China Mountains to the south, and Bohai Bay to the east (Figure 1b). The northern and western regions of the Haihe River Basin were predominantly mountainous, comprising 62.8% of the total area, while the plains to the east and partial south made up the remaining 37.2%. The low-lying eastern plains, facing the ocean, were particularly vulnerable to warm, humid air masses from the southeast during the summer, often resulting in frequent flooding. Rainstorm patterns in the basin were significantly influenced by the terrain, especially along the windward slopes of the Taihang and Yanshan Mountains (Figure 1b). Major rivers, such as the Dongyang, Yinma, and Huangshui, primarily flowed through these mountainous areas, where steep gradients and short runoff times resulted in high runoff coefficients during heavy rainstorms, increasing the risk of flooding. In contrast, during autumn and winter months, cold air masses from the northwest moved across the mountains, transforming into dry, warm currents. This reduced precipitation in the downstream plains and exacerbated drought conditions, making the region prone to both flood and drought variability throughout the year [21,23,24].

5.1.2. Climate Change

Over the past centuries, China has experienced a warming trend consistent with global climate patterns [6,34,35] (Figure 8). This warming was reflected in the drought levels in the northern Haihe River Basin, which aligned with the regional temperature variations and changes in the East Asian monsoon [36] (Figure 8a,b). Between 1905 and 2001, ground temperatures in China rose by an average of 0.8 °C, slightly higher than the global average increase of 0.74 °C. From 1956 to 2002, China’s annual average temperature increased by 0.25 °C per decade, with the most pronounced warming occurring in winter and the least in summer.

Our study indicated that the northern Haihe River region has experienced significant climate variability on decadal to centennial scales, with notable drying in episodes in the 17th and 20th centuries and relatively wetter periods in the 16th, 18th, and 19th centuries. These fluctuations were likely tied to larger environmental shifts during the Holocene, as well as regional and even global climate patterns [6,36,44]. Notably, the northern Haihe River Basin has seen a temperature increase of approximately 0.36 °C per decade over the last 50 years, surpassing the national average (Figure 8b). Winter temperatures showed the most significant rise, while summer temperatures saw the smallest increase. The trend indicated continued warming since the 1990s. Zhai et al. noted that rising ground temperatures boost surface evaporation, raising the risk of drought risks and more frequent floods [36,45,46]. To compensate for this, precipitation also rose, consequently enhancing the frequency of floods. Precipitation trends remained relatively stable but followed a cyclical pattern: a decline from 1890 to 1925, an increase from 1926 to 1960, and a subsequent decline. This cyclical behavior suggests that precipitation might increase again in the coming decades.

The results of wavelet and power spectrum analyses revealed that drought and flood periods in the northern Haihe River Basin from 1470 to 2009 followed periodic fluctuations, with a dominant cycle of 30–50 years (Figure 5 and Figure 6). These climate patterns in the region were closely aligned with broader regional and even national climate trends, indicating that large-scale atmospheric and oceanic influences contributed to shaping the local climate [23,36,43,46]. While the variations observed in the region were less dramatic than major climate shifts, they still resulted in significant changes in temperature and irregular weather fluctuations. Such anomalies, including prolonged periods of excessive rainfall or drought, created favorable conditions conductive to natural disasters like floods, droughts, and other extreme weather events [6,23]. Consequently, even relatively moderate shifts in climate conditions could disrupt local ecosystems, agriculture, and water resources, leading to devastating impacts on the population and economy of the region. This underscores the critical need to understand these periodic climate changes in order to better predict and mitigate the effects of future climatic disruptions.

5.1.3. Solar Activity

Fluctuations in solar activity, particularly related to sunspot cycles, have been a key focus for meteorologists studying their influence on weather and climate [38,39,40]. Sunspots, which served as indicators of solar activity, define the peak and trough years of an 11-year cycle. Research has shown that peak years (M) and the first two years of the solar cycle were associated with notable precipitation variability [47,48], whereas trough years (m) saw fewer large-scale precipitation anomalies. This study examined the correlation between sunspot activity and drought/flood events in the northern Hebei River Basin using sunspot data from the Royal Observatory of Belgium (1755–2009). A 5-year sliding trend analysis of drought and flood levels (Figure 8c), revealed that between 1776–1790 and 1805–1860, sunspot activity trends mirrored drought/flood levels. Peaks in sunspot activity (e.g., 1778, 1830) were followed by corresponding peaks in drought/flood levels after a 2–3-year delay. In most other periods, drought and flood levels inversely correlated with sunspot numbers, indicating that increased sunspot activity was linked to more frequent flooding, while declining sunspot activity correlated with more severe drought periods [23,47]. Between 1755 and 2009, there were 25 peak and trough years. In trough years (m), drought periods occurred in 5 out of 8 years, with significant waterlogging, while in peak years (M), 8 out of 10 years of waterlogging coincided with drought periods. These findings suggested that extreme drought and flood variability were more likely to occur during trough years than during peak years.

5.1.4. El Niño and Southern Oscillation

Sea surface temperature anomalies in the tropical Pacific, particularly those associated with the El Niño–Southern Oscillation (ENSO), significantly influenced atmospheric circulation, affecting East Asian monsoon behavior and contributing to drought and flood periods in regions like the northern Haihe River Basin [6,23,26,49,50]. El Niño events, marked by elevated sea surface temperatures in the equatorial Pacific, were linked to specific atmospheric changes that can disrupt weather patterns, often leading to drought conditions. In contrast, La Niña events, characterized by cooler-than-average sea surface temperatures, created contrasting atmospheric conditions, typically resulting in increased rainfall and a higher likelihood of flooding.

Research has shown a clear correlation between ENSO phases and drought/flood patterns in the northern Haihe Basin (Figure 9). Positive ENSO, which corresponded to El Niño events, was linked to drought years, while negative ENSO values, associated with La Niña events, were connected to flood periods. Of the 125 years analyzed, 58 were associated with drought years and 67 with flood periods, yielding an 81% correlation with ENSO events. During flood years, the probability of an El Niño event occurring was 17%, while during drought years, this probability increased to 37%. Major floods were frequently tied to La Niña events, while significant drought variations were typically linked to El Niño events. These findings highlight a clear trend in the northern Haihe River Basin, with a higher likelihood of droughts during El Niño events and an increased risk of floods during La Niña events [49,50].

5.1.5. East Asian Summer Monsoon and Western Pacific Subtropical High Anomalies

The northern Haihe River system is located within the East Asian monsoon climate zone and is significantly influenced by the East Asian summer monsoon, which shifts rain bands seasonally. Initially, these rain bands were positioned south of the Yangtze River Basin but gradually moved northward by mid-summer, creating contrasting patterns of drought and flood between the northern and southern regions [16,20,39,40,51,52]. Research has highlighted abrupt shifts in the East Asian summer monsoon circulation during early June, often resulting in abnormal precipitation patterns [20,39,40,51,52] (Figure 8d). Wang et al. studied the rapid transitions between drought and flood periods during peak flood seasons in the northern Haihe River Basin, linking these fluctuations to anomalies in atmospheric circulation [53]. Li et al. identified a 40-year cyclical consistency between the behavior of the East Asian summer monsoon and regional hydrological factors, noting a downward trend in hydrological elements over the past 60 years, which corresponded with a weakening monsoon [54]. Looking forward, the East Asian summer monsoon has been expected to experience a significant resurgence around the 2040s, potentially bringing substantial rainfall to the northern Haihe River Basin.

5.2. Anthropogenic Impacts

5.2.1. Population Growth and Expansion of Arable Land

The rapid population growth in the northern Haihe River region has significantly intensified competition for resources, placing considerable strain on the environment. Since 1470, the population has surged, leading to an increased demand for food and agricultural land. For example, Tianjin’s population grew from 1.343 million in 1776 to 4.745 million by 1953 [23,36,41] (Figure 8f). This population boom spurred widespread land cultivation, often at the expense of forests and water bodies. As agricultural land became scarcer, practices such as deforestation and land reclamation became more prevalent. The destruction of lakes, which were critical for water storage and maintaining ecological balance, further exacerbated the frequency of drought and flood variations. Between 1820 and 1953, Tianjin experienced drought and flood periods 70% of the time, a significant increase compared to the 54% frequency observed from 1776 to 1820. This trend highlights the negative impact of population growth and agricultural expansion on the region’s climate and hydrological stability.

5.2.2. Depletion of Forest Resources

Forests, often called the green guardians of the Earth, have long played a crucial role in water conservation and climate regulation. However, extensive deforestation, particularly in the Taihang Mountains during the Ming and Qing dynasties, has led to severe ecological imbalances. The loss of forest cover has diminished the land’s ability to absorb rainfall, resulting in more frequent and intense downstream flooding. Historically, the region was rich in vegetation, but uncontrolled logging for construction and other purposes has significantly depleted these forest resources. This deforestation has accelerated soil erosion, land degradation, and reduced water storage capacity, all of which have heightened the region’s vulnerability to both droughts and floods [13,23].

5.2.3. Water Resource Utilization

The economic development of the northern Haihe River region, particularly in urban centers such as Beijing and Tianjin, has driven a substantial increase in water consumption across industrial, agricultural, and domestic sectors. Between 2001 and 2009, the average annual water consumption in Hebei nearly quadrupled compared to the 1960–1969 period, while water reserves declined by approximately 2 billion cubic meters. Over the past five decades, the region has experienced persistent droughts, compounded by a significant reduction in inflow to major reservoirs like the Beijing Guanting Reservoir, which saw inflows drop to just 40% of the 1990s average between 1999 and 2004. Excessive groundwater extraction has further worsened the situation, with some areas extracting over 150% of sustainable levels, leading to cumulative over-exploitation exceeding 100 billion cubic meters [23,36,55].

Ground subsidence, especially in urban areas such as Tianjin and Tanggu, has become a pressing issue, with subsidence rates reaching up to 2.78 m. This has significantly increased flood risks during rainy seasons. Additionally, water pollution, driven by rapid urbanization, has further exacerbated the situation, threatening drinking water safety and intensifying competition for already scarce resources. This, in turn, has worsened drought conditions and heightened the risk of flooding, especially as algal blooms obstruct waterways.

In conclusion, human activities—driven by population growth, urbanization, and industrialization—have severely disrupted the ecological balance in the northern Haihe River area, contributing to an increase in both the frequency and severity of droughts and floods [6,23].

5.3. Regional Comparison

5.3.1. Comparison with Climate Change in the Haihe River Basin

The Haihe River Basin, encompassing regions such as the Luan River, northern Haihe River, southern Haihe River, and Tuhaimajia River, exhibits distinct climatic variations. This section compared the climatic changes in the Haihe River Basin, with a particular focus on the southern Haihe River (Figure 8a), in contrast to the more limited studies on the Luan River and Tuhaimajia River.

Over the past 500 years, climate changes associated with drought and flood variations in the Haihe River Basin can be divided into three main phases [6,14,17,23]. Phase 1 (before the 18th century): This period was primarily characterized by drought period and a warm, dry climate, with significant dry spells between 1490–1552 and 1608–1652, highlighting prolonged water scarcity. Phase 2 (after the 18th century): The climate shifted toward increased rainfall and cooler conditions, with notable flooding events between 1750–1835 and 1891–1928, signaling a transition to wetter conditions. Phase 3 (since 1949): In this phase, drought conditions reemerged, accompanied by overall warming and drying trends. These recurring dry and wet cycles underscore the region’s climatic volatility [6,14,17,21,23].

A comparison of drought and flood trends between the northern and southern Haihe River provided valuable insights. Shi’s detailed yearly assessment of drought and flood levels across 11 stations in the southern Haihe River from 1470 to 1920 [13] revealed a gradual drying trend over the last five centuries. This trend was marked by three major drought periods: 1581–1618, 1636–1660, and 1680–1731. After 1735, the climate shifted to more humid conditions, with notable wet periods from 1746 to 1810 and a significant wet phase from 1870 to 1920. This comparison indicated that while both the northern and southern Haihe Rivers have experienced similar climatic fluctuations, the timing and intensity of drought and flood periods have differed due to complex regional climatic and human influences. Understanding these dynamics is essential for effective water resource management and disaster preparedness.

Further investigation by Yang and Yao into drought and flood fluctuations in the Hebei Plain, focusing on the Hutuo and Ziya River basins from 1500 to 1996 [6,36] (Figure 8a), identified three distinct periods of alternating humid and dry phases between the 16th and 19th centuries in the southern Haihe River. However, the 21st century has brought a noticeable shift: the northern Haihe River is becoming drier, while the southern region is experiencing relatively wetter conditions. Precipitation trends between 1960 and 1995, shown in Figure 7, highlighted a consistent increase in rainfall in the southern Haihe River, accompanied by significant flooding trends post-1994. Other studies confirmed that annual precipitation has increased more significantly in the southern Haihe River than in the northern regions [56,57]. Recent data revealed a maximum precipitation of 754.3 mm in the south compared to 675.9 mm in the north, with minimum levels of 396.7 mm and 370.0 mm, respectively [58].

The southern Haihe River Basin, including the Daqing, Hutuo, and Ziya Rivers, is larger and has more tributaries, contributing to higher moisture levels. In contrast, urbanization in the northern region, particularly in cities like Beijing and Tianjin, has significantly increased the demand for water, leading to a gradual decline in river and lake volumes [14,23]. This reduction has diminished the capacity of these water bodies to regulate the climate, exacerbating dryness in the northern Haihe region. Consequently, while the southern Haihe area has experienced increasing precipitation, the northern areas have faced dwindling water resources, revealing a critical imbalance with significant implications for regional water management and climate resilience strategies.

5.3.2. Comparison with Climate Change in the Eastern Region

The climate dynamics of the eastern region, particularly in the middle and lower reaches of the Yangtze River, exhibited significant historical drought and flood patterns [23,59]. The driest and wettest years in this region, post-1500, occurred in 1589 and 1954, respectively. In contrast, northern China experienced its driest year in 1640 and its wettest in 1964 [59,60]. Specifically, the northern Haihe River system faced severe drought years in the years 1560–1561, 1640–1641, 1900, 1920, 1972, and 1997, while major floods occurred in 1652–1653, 1801, 1871–1872, 1890–1894, 1917, and 1959. Notably, the driest and wettest periods for the northern Haihe River after 1500 were recorded in 1640 and 1871, respectively.

A comparison of climatic trends indicated that the driest periods in both the eastern region and the northern Haihe River Basin coincided with the Ming and Qing Little Ice Age (1550–1851), while the wettest period in the eastern region aligned with the modern warming phase (Figure 8g,h). This analysis suggested a synchronicity in climate change patterns during the driest and wettest phases between the Haihe River Basin and the eastern region [6,42,43,60,61]. When examining the last 500 years of climate change, both regions exhibited consistent temperature trends. During the Little Ice Age, both regions experienced relatively dry conditions, while the modern warming period has led to a more humid eastern region, in contrast to the drying trend in the northern Haihe River Basin [23,60,61]. This divergence suggested that both regions were influenced by the East Asian monsoon, with alternating patterns of dry and wet as well as warm and cold climates. Additional research indicated a continuous weakening of the summer monsoon since the 1970s [62,63]. A weaker summer monsoon might shift the rain belt southward, which could help explain the trend of aridification in the northern Haihe River system over the past 500 years [60,61,62,63]. This interplay of climatic factors highlighted the complexity of regional climate change and its implications for water resource management in both areas [64].

5.4. Drought–Flood Response Strategies

To enhance drought resistance and optimize agricultural water use, the adoption of advanced irrigation techniques is essential. Farmers should be encouraged to implement efficient methods such as lined irrigation channels, field pipelines, sprinkler systems, and drip irrigation to minimize water wastage. Additionally, developing rainwater utilization technologies can help tackle irrigation challenges during critical dry periods, fostering an integrated water management approach focused on storage, conservation, collection, and efficient use.

It is also important to monitor groundwater levels regularly to prevent secondary soil salinization. In areas where appropriate, irrigation methods that utilize brackish water should be considered. Establishing a comprehensive spatiotemporal monitoring system for agricultural water use will allow the real-time assessment of water requirements and promote sustainable practices.

Creating a robust drought–flood monitoring and prediction framework is essential. This includes developing effective data collection and analysis systems, as well as an information-sharing platform to facilitate clear communication about drought conditions. A strong legal and regulatory framework for disaster prevention will ensure collaboration between government agencies and communities in responding to drought situations.

Aligning regional social and economic development strategies with available water resources is also beneficial. Key factors such as GDP, population, and economic structure should be considered when planning to mitigate the effects of droughts. In terms of ecological management, it is important to balance water resource management with environmental protection, flood control, irrigation, and drainage, thereby creating a cohesive regulatory framework.

Additionally, the development of a water reserve system and emergency response protocols for severe drought conditions is essential. These measures will ensure the rapid allocation and distribution of water resources when necessary. Enhancing soil and water conservation practices is also critical because reducing soil erosion can help restore ecological balance, alleviate drought impacts, and reduce the frequency of drought-related disasters.

Lastly, promoting a water-saving culture is key. Public education on disaster preparedness and water conservation should be prioritized to emphasize the importance of ecological protection alongside agricultural production. Various media channels can be leveraged to encourage community engagement in drought prevention initiatives, fostering a sense of collaboration. Technical personnel should engage with local communities to apply drought-resistant and water-saving technologies while guiding farmers on best practices, reinforcing a water-efficient agricultural system.

By implementing these strategies, the region can improve its drought resilience, optimize water resource management, and promote sustainable agricultural practices. This is of great importance for accelerating the development of a new growth model, advancing China’s water-saving initiatives, promoting comprehensive water resource management, and ensuring a safe and equitable water future.

6. Conclusions

This study analyzed drought and flood data from seven stations in the northern Haihe River Basin over the past 500 years, utilizing the watershed drought and flood level method. It reconstructed drought and flood sequences from 1479 to 2009 using analytical methods like sliding averages, the Mann–Kendall test, continuous wavelet analysis, and ArcGIS, while examining the impacts of climate change, ENSO, sunspot activity, and human influence. The findings reveal that drought and flood periods occurred 73.15% of the time, with an increasing trend. Flood periods were predominant before the 20th century, while drought variability became more frequent since post-1950. Drought and flood cycles have become more variable with shorter oscillation periods. The primary driver was the uneven distribution of precipitation, influenced by the monsoon climate, ENSO events, sunspot activity, and human activities like deforestation. Looking forward, the northern Haihe River Basin has been expected to face more aridification and water supply challenges over the next 40 years. This study suggested strategies to mitigate these impacts, such as promoting water-saving agricultural technologies, improving water use efficiency, and enhancing soil and water conservation efforts. Climate patterns in the region have alternated between dry and wet periods, with recent decades showing a shift toward increasing dryness. This study examined historical documents over 540 years and used various mathematical methods to offer quantitative insights for future drought and flood research. However, the analysis of the factors driving drought and flood formation is limited, and the interactions within the global climate system—particularly the underlying mechanisms—need further exploration. Further research should focus on targeted strategies for managing the evolving challenges of drought and flood events.