Next Article in Journal
Flood Loss Models and Risk Analysis for Private Households in Can Tho City, Vietnam
Next Article in Special Issue
Current Assessment and Future Outlook for Water Resources Considering Climate Change and a Population Burst: A Case Study of Ciliwung River, Jakarta City, Indonesia
Previous Article in Journal
Effects of Precipitation and Topography on Total Phosphorus Loss from Purple Soil
Previous Article in Special Issue
Sustainable Measures for Mitigation of Flooding Hazards: A Case Study in Shanghai, China
Article Menu
Issue 5 (May) cover image

Export Article

Water 2017, 9(5), 316; doi:10.3390/w9050316

Article
Changes of Reference Evapotranspiration and Its Relationship to Dry/Wet Conditions Based on the Aridity Index in the Songnen Grassland, Northeast China
Qiyun Ma 1, Jiquan Zhang 1,*, Caiyun Sun 1, Enliang Guo 1, Feng Zhang 1 and Mengmeng Wang 2
1
Institute of Natural Disaster Research, School of Environment, Northeast Normal University, Changchun 130024, China
2
College of Resources and Environment, Northeast Agricultural University, Harbin 150030, China
*
Correspondence: Tel.: +86-135-9608-6467
Academic Editor: Franco Salerno
Received: 22 February 2017 / Accepted: 26 April 2017 / Published: 29 April 2017

Abstract

:
Reference evapotranspiration (ET0) plays an important role in regional dry/wet conditions. Based on the Food and Agriculture Organization of the United (FAO) Penman-Monteith method and daily climate variables, ET0 was calculated for 21 stations in and around the Songnen Grassland, northeast China, during 1960–2014. The temporal and spatial variations of ET0 and precipitation (P) were analyzed in the annual, seasonal, and growing season (from April to October) time series using the Mann-Kendall test, Sen’s slope estimator, and linear regression coupled with a break trend analysis. A sensitivity analysis was used to detect the key climate parameter contributing to ET0 change. By linear regression analysis on the relationship between ET0, P, and the aridity index (AI), the role of ET0 in determining regional wet/dry conditions was analyzed. Results show a higher ET0 in the southwest and a lower ET0 in the northeast, but P was opposite to that of ET0. Evident decreasing trends of ET0 in the annual, seasonal, and growing season time series were detected in almost the entire region by the trend analysis methods. For the entire region, the decreasing trend of ET0 can be linked to the relative humidity and maximum air temperature. The positive contribution of increasing temperature to ET0 was offset by the effect of the significantly decreasing relative humidity, wind speed, and sunshine duration at the 0.05 level during 1960–2014. In addition, the value of ET0 was higher in drought years and lower in wet years.
Keywords:
reference evapotranspiration; climatic change; drought/wet; Songnen Grassland

1. Introduction

Climate change is an indisputable fact and may accelerate hydrological cycles and redistribute global water resources [1,2]. The change has been identified not only in individual parameters, such as temperature or precipitation, but also in integrated parameters, like reference evapotranspiration (ET0) [3]. ET0 is one of the vital components of the hydrological cycle and controls the energy and mass exchange between terrestrial ecosystems and the atmosphere [3,4]. It is influenced by many factors including climate factors (e.g., temperature, solar radiation), crop factors (e.g., crop pattern, cropping system), environmental conditions, and water resource management [5]. Changes in ET0 affect agricultural production, water resource programming, and irrigation systems. The observation of trends in these climate factors can reveal the possible impacts of climate change or climate natural variability on the hydrological cycle, identify the spatial and temporal variation, and determine the dominant climatic variables affecting ET0. These trends can be helpful in determining appropriate measures for mitigating the potential damage from climate change or its natural variability effects [6,7,8]. However, the relationship between changing ET0 and the dry/wet tendency is not quite clear, though it is crucial for water resource management.
Studies that investigate changes in climate factors have yielded a mixture of results and conclusions about the trends of ET0 for specific locations during the last few decades. Contrary to the general expectation that an increase in temperature leads to an increase in evapotranspiration, such as in south and southeast Romanian during 1961–2007 and in the semi-arid and humid regions in Iran during 1966–2005 [3,9], some previous studies conclude that evaporation has diminished in the last few decades. In the upper and middle to lower Yangtze River basin, Wang et al. [10] reported a decreasing trend in ET0 from 1961–2000 based on daily data from 115 meteorological stations. Over a 116-year period (1983–2008), Irmak et al. [11] found a significant decrease in ET0 in the Platte river basin, USA. The same declining trends were also found in other regions throughout the world, such as New Zealand (from the 1970s to the early 2000s), India (1971–2002), and parts of China (1960–2010) [12,13,14]. However, an increasing trend in ET0 was identified in other regions from the 1960s to early 2000s, such as Iran (1965–2005), Northern Eurasia (1948–2005), and parts of Romania (1961–2007) [3,15,16].
Being a typical farming-pastoral ecotone located in the central part of northeast China, the Songnen Grasslands have experienced high spatial and temporal variability in ET0 and precipitation. This makes the management of water resources difficult in the region. However, to the best of our knowledge, there has been no comprehensive study of the relationships between changes in ET0 and dry/wet conditions, and especially, the sensitivity of ET0 to climatic variables in this region. Water is the lifeline for socioeconomic development in the Songnen Grassland, because agriculture and animal husbandry, which heavily depend on precipitation and irrigation, are the main industries in the area. Therefore, understanding ET0 trends and their role in regional dry/wet conditions could give a scientific basis for regional water resource management and allocation and for the scientific decision-making of preventing flood and drought disaster.
To address the problem of the unclear ET0 changes in the Songnen Grassland, this study set the following objectives: (1) to evaluate the spatial distribution of the ET0 and precipitation at the annual, seasonal, and growing season time scales in the Songnen Grassland over the period from 1960–2014; (2) to investigate the temporal trend of ET0 by using the Mann-Kendall test, and linear regression coupled with a break trend analysis, and the slopes of trend lines using the Sen’s slope estimator; (3) to analyze temporal trends in the climatic parameters needed to calculate ET0 and the sensitivity and trends of ET0 related to the climatic parameters using the sensitivity analysis method; and (4) to explore the role of changing ET0 in regional dry or wet conditions based on the aridity index (AI). The results of this study will improve our understanding of the impact of climate change or climate natural variability on regional hydrological processes and agricultural irrigation management.

2. Materials and Methods

2.1. Study Area

The Songnen Grassland is located in central northeast China. It lies from 43°30′ N to 48°05′ N and from 122°12′ E to 126°20′ E, and covers an area of approximately 22,350 km2 (Figure 1). Generally, the study area is located in the meadow steppe belt of China and is an important grassland in the Eurasian steppe zone. The region is dominated by a temperate continental monsoon climate, with four distinct seasons: quite dry (drought in nine years out of ten), windy springs; warm, rainy summers; sunny, mild autumns; and long, freezing, and dry winters [17,18]. From 1960 to 2014, the mean annual temperature ranged from 1.9 °C in the southwest to 6.2 °C in the northeast, while the mean annual precipitation amount varied from 350 mm in the southwest to 500 mm in the northeast. Meanwhile, the mean annual amount of evaporation is roughly two or three times the precipitation.
The Songnen Grassland forms a typical agricultural area in the east and an agro-pastoral transition zone in the west, determined by various physical geographical features and regional climatic differences [17]. The current ecological environment tends to be deteriorating (e.g., grassland degradation, desertification, and alkalization) due to recent and ongoing climate change and land use change in the region. Since agriculture and animal husbandry development are the main priorities within the general economic strategy of the region, an investigation of changes in ET0 and its role in regional dry/wet conditions are needed for agriculture managers and stakeholders.

2.2. Climate Data and Quality Control

Climate data from 21 meteorological observatory stations were provided by the National Meteorological Information Centre of China, including daily observations of maximum air temperature (Max T, °C), minimum air temperature (Min T, °C), average air temperature (Ave T, °C), average relative humidity (Ave RH, %), wind speed (Win S, m/s), sunshine duration (Sun H, h), and precipitation (P, mm), for the period from 1960–2014. The regional values for the seasonal, growing season (from April to October), and annual climatic variables were then calculated by the weighted average or sum method. The weight of every station was obtained by the Thiessen polygon method, which assigns the weight in accordance with the areas of each station [19]. Thermal seasons were considered as winter (December, January, and February), spring (March, April, and May), summer (June, July, and August), and autumn (September, October, and November).
The location of the chosen meteorological stations is shown in Figure 1. They are all distributed inside or adjacent to the study area in order to cover the entire region. All selected stations have good-quality data and meet the QA (Quality assurance)/QC (Quality control) requirements, which means that the spatial distribution of the stations had to cover the entire region, the time series had to be long enough to obtain statistically significant results in trend analyses, and the climate data was quality controlled by the China Meteorological Administration (http://cdc.cma.gov.cn). The missing data were substituted with the corresponding long-term mean value. Although the infilling of missing data may have a slight influence on the trend analysis for each climatic variable and the computation of ET0, the effect may be negligible since the missing data are infrequent and account for only 0.05–0.1% of the data.

2.3. Methods

2.3.1. Calculation of Reference Evapotranspiration (ET0)

The FAO-56 Penman-Monteith (FAO-PM) equation was recommended as the sole and global standard method for ET0 calculation [5,20]. The method has been widely verified for its accuracy and reliability in various climatologic zones around the world [21,22]. Accordingly, the FAO-PM method was used to estimate the daily ET0 in this study; and, subsequently the seasonal, growing season, and annual ET0 values were derived from the daily values. The FAO-PM equation is expressed as:
E T 0 = 0.408 Δ ( R n G ) + γ 900 ( T + 273 ) U 2 ( e s e a ) Δ + γ ( 1 + 0.34 U 2 )
where ET0 is the reference evapotranspiration (mm/d), Δ is the slope of the saturation vapor pressure curve at a given air temperature (kPa/°C), R n is the net radiation at the crop surface (MJ/(m2·d)), G is the soil heat flux density (MJ/(m2·d)), γ is the psychrometric constant (kPa/°C),   T is the mean daily air temperature at 2 m height (°C), U 2 is the wind speed at 2 m height (m/s), e s is the saturation vapour pressure (kPa), e a is the actual vapour pressure (kPa), and ( e s e a ) is the saturation vapour pressure deficit (kPa).
R n is the difference between the incoming net shortwave radiation ( R n s ) and the net outgoing longwave ( R n l ) radiation. R n s is calculated as
R n s = ( 1 λ ) R s
where R s is the incoming solar radiation (MJ/(m2·d)) and λ (= 0.23) is the albedo of the hypothetical grass reference crop (dimensionless). R s was estimated based on the sunshine duration record according to the calibration equation by Croitoru et al. [3].
At the same time, R n l is given by
R n l = σ [ T max , K 4 + T min , K 4 2 ] ( 0.34 0.14 e a ) ( 1.35 R s R s o 0.35 )
where σ is the Stefan-Boltzmann constant (= 4.903 × 109 MJ/(K4·m2·d)); T m a x , K is the maximum absolute temperature during the 24-h period (K = °C + 273.15); T m i n , K is the minimum absolute temperature during the 24-h period (K = °C + 273.15); and R s o is the clear-sky radiation (MJ/(m2·d)). All of the variables in Equation (1) were calculated using the standard procedure outlined by Allen et al. [20]. The monthly time series of ET0 during 1960–2014 were obtained based on the calculation of ET0 using the CROPWAT 8.0 software developed by FAO (http://www.fao.org/land-water/databases-and-software/cropwat).

2.3.2. Trend Analysis

To calculate trends, six data sets, including one annual, four seasonal, and one growing season, were obtained for each meteorological station. In this study, the Mann-Kendall (MK) test, a non-parametric test recommended by the World Meteorological Organization (WMO), was used to detect trends [23,24,25]. The MK test has been widely used for trend detection in hydrological and climatic research since the data does not need to conform to any distribution form, and it allows the missing data [26,27]. In this research, the missing data were filled due to computation of the ET0. The change points were detected based on the sequential MK test, which has been largely employed for the detection of change points in different climatic and hydrological data series [28,29,30]. Detailed information related to the sequential MK test can be found in studies published by Palizdan et al. [29] and Liang et al. [30]. In addition, Sen’s slope estimator was used to measure the magnitude of the trend [31]. The MK test and Sen’s slope estimator calculations for various time series of climate factors and ET0 were performed using the Excel-based template MAKESENS 2.0 beta, developed by researchers at the Finnish Meteorological Institute [32].
Linear regression coupled with break trend analysis [33] was applied to calculate the temporal trends of the ET0 series for the period from 1960–2014. This method is one of the linear trend methods, which divides a certain time series into several periods. The linear regression is used in each period for detecting the temporal trend. A break trend analysis can provide more details about a long term change pattern compared with a simple linear regression. The period for the break trend analysis was chosen as every 30 years, which was considered as the climatological normal period according to the recommendation of the WMO [34]. In this study, the climatological normal ranges were from 1960 to 1989, from 1970 to 1999, from 1980 to 2009, and from 1985 to 2014. It should be noted that in both trend analysis methods, the significance level ( α ) was set to 0.05 (one-sided t-test).

2.3.3. Sensitivity Analysis

A sensitivity analysis is a quantitative description method for the effect of the input variables on the output [35]. In the present study, it was performed to evaluate the effect of climatic variables on ET0. Because of the different approaches used in determining ET0, there are no standard or common procedures for carrying out the sensitivity analyses on ET0 [20,36]. This study applied the sensitivity analysis method developed from Zhan et al. [37], which has the advantages of simple procedures and evident outcomes of detecting sensitive factors. The measure S x is the sensitivity of the FAO-PM method to a meteorological parameter, defined as:
S x i j = | E T 0 1.1 x i j E T 0 0.9 x i j E T 0 x i j |
where x i is a meteorological parameter needed for the calculation of ET0; j is a given year; E T 0 x i j ,   E T 0 1.1 x i j , and E T 0 0.9 x i j are the estimated ET0 in j year when the parameter x i equals its value or is 1.1 and 0.9 times its value, respectively. The larger the S x i j is, the more sensitivity x i will have.

2.3.4. Aridity Index (AI Index)

Aridity is usually expressed as a comprehensive function of precipitation, temperature, and/or potential evapotranspiration, and reflects the level of meteorological drought [38]. AI is defined by Thornthwaite [39], who calculated AI as dividing the difference between precipitation and potential evapotranspiration by the potential evapotranspiration. This method is understood as the dearth of water availability at the surface and subsurface levels, can express the aridity degree in arid or semiarid regions, and can help to further analyze drought-wetness variations in terms of agriculture demand [4,40].
In this study, following Thornthwaite [39], the AI was computed as:
A I = ( E T 0 P ) / E T 0
where ET0 is the reference evapotranspiration, and P is the precipitation. If the AI is equal to or close to 1, it indicates that there is no precipitation, and aridity is the highest. In contrast, if the precipitation is equal to or higher than ET0, the AI will be equal to 0 or negative. The AI values of the growing season at each station were calculated, and the average AI in the entire Songnen Grassland was expressed by the arithmetic average from the stations in and around the region.

2.3.5. Spatial Interpolation

For analyzing the spatial patterns of ET0 trends and their magnitudes, the Inverse Distance Weighting algorithm (IDW) was used, which is extensively applied for mapping the spatial extent of climatic and hydrological point data. This method is a simple deterministic interpolation, and assumes that sample values closer to the prediction location are more representative of the prediction value than sample values farther away [41]. Therefore, the closest value to the prediction location gets the greatest weight and the weight is reduced as a function of distance [29]. The power parameter ( α ), which controls the importance of the predicted location values, is the main factor affecting the result from the IDW method. The α , a positive, real, and constant number, can change from 1 to 5. As a higher α value is selected, more weight is given to the closed location. In this study, α was defined as 2 by following Palizdan et al. [29]. All spatial interpolations were conducted using the ArcGIS 10.2 software.

3. Results

3.1. Spatial Distribution of ET0, P, and Their Difference over the Period from 1960–2014

The spatial distributions of ET0 and P annually, seasonally, and during the growing season from 1960 to 2014 are shown in Figure 2, and the difference between the two is displayed using the IDW interpolation as well. Figure 2 shows a strong variability and marked difference between the northeastern ranges and the southwestern region. From the visual inspection, it was observed that the spatial pattern of the high and low values of annual ET0, P, and their difference all show similarities to the growing season and seasonal time series. The spatial pattern of ET0 and the difference indicated higher values in the southwest and lower values in the northeast areas, but precipitation showed the opposite trend, in which the minimum value was in the west and the highest value was in the east. Generally, the ET0 value was more than double the precipitation amounts for all time scales considered, especially for the annual values and during the growing season.
For ET0, the average of the annual and growing season values increased from the northeast to southwest, with values less than 850 and 750 mm, respectively, to more than 1100 and 980 mm, respectively. Seasonally, all of the stations showed higher ET0 in the summer, followed by spring, autumn, and winter. For P, the annual value and that of the growing season increased from west to east, ranging from 381 to 577 mm and from 352 to 511 mm, respectively. The minimum was found in Tongyu, and the maximum was found in Changchun. Seasonally, higher amounts of precipitation were exhibited during the summer, followed by autumn, spring, and winter. The maximum difference was distributed in the southwest at all time scales (Figure 2), especially for the annual values, reaching 700–800 mm in the southwest. From a seasonal perspective, the maximum value of the difference occurred in the spring, ranging from 215 to 350 mm and was followed by summer, autumn, and winter.

3.2. Temporal Variations of ET0

The results of the break trend analysis for the area-averaged annual, seasonal, and the growing season ET0 time series of the 21 stations during 1960–2014 are shown in Figure 3, and those of the MK test and Sen’s slope estimator are summarized in Figure 4. Figure 3 indicates an evident decreasing trend in ET0 in the annual, seasonal, and growing season time series, especially for the annual, growing season, and spring periods, which passed the significance test at the 0.05 level. In addition, it was shown that almost all annual, seasonal, and the growing season time series had experienced a decreasing trend of ET0 for the four climatological normal time ranges with the rate ranging from −2.415 to −0.003 mm/a. The sharpest drop rate was detected from 1970 and 1999, and the maximum was in the growing season (−2.415 mm/a). However, it is noteworthy that the region experienced an increasing trend of ET0 in the annual, seasonal, and the growing season time series during 1980–2009, with a rate ranging from 0.085 to 0.296 mm/a, except in spring and winter, although the increasing trend was not significant. To detect the change points, the MK test was performed in the ET0 for the annual, seasonal, and the growing season time series during 1960–2014. As shown in Figure 3, both the annual and growing season series have three change points, 1989, 1994, and 2009, and 1989, 1995, and 2008, respectively. Seasonally, in spring, there was a clear change point in 2008; in autumn, four change points were identified, 1972, 1985, 1994, and 2009; in winter, the series had change points at 1977 and 2011. However, there was no detectable change point in summer. The results show that abrupt changes in the ET0 series mainly occurred in the early and mid-1990s, leading to a fluctuation of ET0 which is dominated by a decreasing trend.
The results of the MK test are exhibited in Figure 4, and they are almost identical to those from the linear trend analysis. ET0 exhibited a decreasing trend in the annual, seasonal, and the growing season time series at the 21 stations during 1960–2014. Overall, more than 90% of the analyzed stations reported decreasing trends of ET0 except during autumn and winter, and among those stations about 50% were shown to be decreasing during the annual, growing season, and spring in a statistically significant manner. Increasing trends were found in less than 10% of the data sets, especially during autumn with 38% showing increasing trends, but none of the trends were significant. In terms of the magnitude of the trends, the highest decrease in ET0 was recorded in Haerbin (around the region) and Fuyu City (inside the region) at the annual level, with an average decrease of −4.45 and −2.86 mm/a, respectively. On the basis of these results, there is no doubt that the Songnen Grassland has experienced an evident decrease in ET0 over the last 55 years. With decreasing ET0, the crop or vegetation water demand would be reduced, and the risk of drought would be mitigated. In other words, the decrease in ET0 rates may mitigate the drought impact on local vegetation and agricultural production.

3.3. Spatial Patterns of Trends in ET0

The spatial variations of ET0 trends at different time scales from 1960 to 2014 based on the MK test are displayed in Figure 5. All considered time scales indicated a similar pattern, with the majority of the time series of ET0 indicating a downward trend in the entire region, except during autumn. The greatest number of stations with significant decreasing trends of ET0 were distributed in the eastern, northeastern, and central regions and frequently occurred in the annual, spring, and growing season periods. These regions also had the highest rate decrease in ET0, ranging from −0.2 to −4.0 mm/a. However, positive changes were recorded in the southwestern and southern regions in autumn and winter, but they were statistically insignificant, with rate increases ranging from 0.01 to 0.40 mm/a. Overall, the spatial pattern of ET0 change in the Songnen Grassland is that of a significant downward trend across the central region with a gradually reduced intensity from the eastern parts to the western areas.

3.4. Changes in the Climatic Parameters

The changes in basic climatic parameters that might play a considerable role in ET0 change have been investigated for the 21 stations across the Songnen Grassland during 1960–2014. With an increase of air temperature (Ave T, Max T, and Min T), the regional annual Ave RH, Sun H, and Win S all significantly decreased during the study period (Figure 6). All the climatic parameters experienced constant changes during the four climatological normal periods, i.e. from 1960 to 1989, from 1970 to 1999, from 1980 to 2009, and from 1985 to 2014. All the increasing trends of Ave T, Max T, and Min T and the decreasing trends of Sun H and Win S passed the significance test at the 0.01 level, while the Ave RH passed the significance test at the 0.05 level. The MK test was used to detect the change points for each climatic parameter. Each of them has a clear change point around 1990. The change point years of Ave T, Max T, and Min T with significant increasing trends were 1987, 1989, and 1983, respectively. Those of Ave RH, Win S, and Sun H were 1993, 1990, and 1987, respectively. It is worth noting that the time of abrupt changes in Max T and Ave RH are almost in line with that of the annual ET0 time series during 1960–2014. In addition, the detected change points for both the annual ET0 and the climate parameters mainly occurred in the late 1980s and 1990s.
The MK test was used for detecting the statistical significance of the climatic parameters trends at the 21 stations in the annual, seasonal, and the growing season time series during 1960–2014. The results are summarized in Figure 7. For Ave T and Min T, the percentage value of the statistically significant increase (α = 0.05) was 100% at all considered time scales, except during winter (95% in Min T and more than 70% in Ave T). In the case of Sun H, all of the annual, seasonal, and growing season series illustrated downward trends, ranging from 52% to 71% and being statistically significant decreases. As for Win S, significantly decreasing trends were found roughly 100% of the time at all locations. For Ave RH, an overwhelming majority of data series saw negative trends, but in spring, positive trends were detected at more than 50% of the stations.

3.5. Sensitivity Analysis of Climatic Variables

The results of the sensitivity analysis for annual ET0 to the climatic variables are given in Figure 8 and Table 1 for the 13 stations inside the Songnen Grassland. It can be seen from Figure 8 that the Ave RH and Max T were the most sensitive variables to the change in ET0 over the whole region, and the sensitivity decrease from Win S, Sun H, Min T to Ave T. However, there was a small difference in the northeast region at the Keshan and Mingshui stations, which may be the result of the relatively high elevation of the two stations, and there are small hills near those stations. The most sensitive climatic variables at these two stations were Ave T and Min T, respectively. The summary of the sensitive climatic variables and the temporal trends of their sensitivities based on the MK test are given in Table 1. The results indicate that the sensitivity of Ave RH had a significant decreasing trend except at the Anda station, which had a significant increasing trend. The trend of the sensitivity of Max T was the same as that of Ave RH with a significant downward trend for almost the whole region, except for a significant upward trend in Changchun. For the sensitivity of Win S and Min T, a majority of stations had a significant increasing trend. Sun H had the same temporal trend as Win S and Min T; but in the southwest region, Tongyu and Changling, there was no significant decreasing trend. As for the sensitivity of Ave T, the results showed a slight impact on the changing ET0 in the Songnen Grassland, except in Keshan. To summarize, the sensitivity of Ave RH, Max T, Ave T, and Min T displayed significant decreasing trends, while that of Win S and Sun H displayed significant increasing trends in the whole region.

3.6. The Role of ET0 in Regional Dry/Wet Conditions

The ratio of ET0 during the growing season to the total annual amount was computed from 1960 to 2014. Results show that, in general, the total ET0 in growing season accounts for more than 80% within the year. As this period is critical for crop and vegetation growth, it directly affects regional socioeconomic development. Therefore, this study only considers the role that ET0 may play in regional dry/wet conditions during the growing season. Regional trends and the magnitudes of ET0, AI, and P were analyzed at two spatial scales, at the 13 inside stations and at the 21 stations which takes the adjacent stations into account, respectively. However, only the results of the 13 interior stations are presented in Figure 8, due to the similarity of the results for the analysis of the trends and the magnitudes of ET0, AI, and P at the 21 stations.
Figure 9 gives the interannual variations of the growing season ET0, AI, and P over the Songnen Grassland for the period from 1960 to 2014. From the visual inspection, ET0, AI, and P all displayed decreasing trends, but only the decreasing trend of ET0 passed the significance test. As for the fluctuation in temporal evolution, the results indicated that the fluctuating pattern of ET0 was similar to AI, but opposite that of P. A simple regression equation was established between ET0, AI, and P based on the SPSS Statistics 19.0 Software. The equation is as follows:
A I = 63.81 0.125 P + 0.046 E T 0   ( R 2 = 0.98 ,   P = 0.00 )
Equation (6) indicates a positive correlation between AI and ET0 and a negative correlation between AI and P. The regression residual standard error is displayed in Figure 10 to test the validity and accuracy of Equation (6). It indicated that the residuals are approximately normally distributed since all 55 residuals are distributed around the diagonal line, and it also illustrated that the regression equation (Equation (6)) is in line with the normal distribution. Thus, Equation (6) is statistically significant. Furthermore, the upward trends of ET0 and AI were associated with the downward trends of P and vice versa. This characteristic was highly explicit during the drought and wet years in particular. As shown in Figure 10, during the severe drought years D1 (1982), D2 (2001), D3 (2004), and D4 (2007), the values of ET0 and AI had reached or were close to the maximum in the last few decades; however, the values of P had almost reached a minimum at the same time. Similarly, the wet years, such as W1 (1960), W2 (1998), and W3 (2013) had lower values of ET0 and AI (Figure 9).

4. Discussion

The Songnen Grassland shows obvious spatial variations of ET0 rates from the northeast to southwest at different time scales (Figure 2). These variations are influenced by different regional climates and local topography. In the northeast and east, the temperature is comparatively low, because there were some valleys and the stations were located in a relatively high latitude area combined with the cold temperate continental monsoon climate, leading to the lower values of ET0 in these areas; while in the southwest, the topography was relatively flat at lower latitudes (Figure 1), in addition to being far away from the ocean, resulting in the lower relative humidity and the higher values of ET0. Moreover, in this study, the IDW method could affect the accuracy of the spatial distribution as well. Previous studies indicated that the interpolation methods should account for external factors, such as altitude and land use patterns [42]. However, compared with other interpolation methods, the IDW produced slightly better predictions of ET0 [29,43]. Future studies should consider how different interpolation methods impact the spatial variations of ET0 at different time scales.
Moreover, the trend analyses of ET0 at different time scales by linear regression and non-parametric tests were consistent. Results show that the ET0 rates over the Songnen Grassland have been decreasing over the last few decades (Figure 3, Figure 4 and Figure 5). The results were consistent with decreasing trends in the lower reaches of the Taoer River basin in Northeast China investigated by Liang et al. [30]. Huo et al. [4] also reported a decreasing trend in the arid area of northwest China during 1955–2008. However, as discussed in the introduction, some studies have identified increasing trends in some regions during the last few decades. One of the reasons for the inconsistent findings in ET0 trends is due to the fact that some studies on climate change utilize divergent climatic parameters, potentially providing incomplete or artificial trends and magnitudes of ET0 [11].
During the study period, the observed variations in Max T, Min T, and Ave T were similar to the global pattern of increasing minimum, mean, and maximum temperature [12]; and the results of the analysis of other climatic factors were consistent with the findings reported from Yunnan Province in southwest China during 1961–2004 [5]. There, the sensitivity analyses for changes in the annual ET0 showed that Ave RH and Max T were the most sensitive climate variables over the whole Songnen Grassland, which was in agreement with Yin et al. [44], who studied all of China during 1961–2008. However, Liu et al. [14] has reported that Win S and Sun H were the most sensitive factors in other parts of China. Therefore, it can be concluded that ET0 has different responses to climate variables in different regions and under different climate conditions. The reasons for the inconsistent responses of ET0 may be the geographical position (e.g., distance from the ocean, the altitude variations) and the research period. In addition, this study indicates that the increase of ET0 induced by rising air temperature can be compensated by the reduced ET0 with the significant decrease of Ave RH, Win S, and Sun H. As a result, the regional ET0 appeared to show a declining trend. Human activity could account for some of the local climate change [36], such as greenhouse gas (GHG) emissions and rapid urbanization, which should be further studied in the Songnen Grassland. In addition, the existence of natural climate variability may also affect the ET0 trends. Further research should focus on how to identify the effect of climate natural variability in the changes of ET0, for better understanding of the mechanism of ET0 variations.
Comparing the results of the detected change points of ET0 with the climatic factors in the annual time scale during 1960–2014, abrupt changes were detected in 1989, 1994, and 2009, while the most sensitive factors, Ave RH and Max T, were detected in 1993 and 1989. There is a good correspondence between these two sensitive factors and the ET0 series. However, undoubtedly, the abrupt change of other climate factors also played an important role in the abrupt changes of ET0. From 1989 to 1994, the ET0 experienced an abrupt decreasing trend due to the abrupt increasing trend of Max T, Min T, and Ave T, and decreasing trend of Sun H and Win S; the Ave RH also experienced an increasing trend until it reached the year of abrupt change in 1993. From 1994 to 2009, the ET0 experienced an abrupt increasing trend arising from the abrupt increasing trend of Max T, Min T, and Ave T, and the decreasing trend of Ave RH, Sun H, and Win S. After 2009, the temporal trend of ET0 decreased again, although no change points of the climate factors were detected. This can be due to, as shown in Figure 3, the decreasing trends of every climate parameter except for the Ave RH.
As a whole, the abrupt changes of ET0 and its related climatic factors mainly occurred in the early and mid-1990s, and it may be concluded that climate variations were more dramatic since this period, when the air temperature remarkably increased, with decreasing trends of Ave RH, Win S, and Sun H. This period of abrupt surface warming was consistent with Dong et al. [45], who reported an abrupt increase in summer mean surface air temperature over Northeast Asia since the mid-1990s. Accompanying the above mentioned climatic anomalies, a decrease in ET0 in the Songnen Grassland was observed. In addition, previous studies indicated that the northeast China climate is affected by both the thermal contrast between the Asian continent and the tropical Indo-Pacific Oceans and that between the continent and the extratropical North Pacific [46]. The western Pacific subtropical high (significant positive correlation with summer temperature at the 0.05 level during 1994–2012), zonal circulation (same as above), and Arctic Oscillation (high positive correlation with winter temperature during 1951–1999) also have strong correlations with regional climate variations over northeast China [47,48]. The temperature is one of the most sensitive factors for the change in ET0, therefore, the large-scale atmospheric and ocean circulation may be other reasons for the abrupt change and variations in ET0. Future studies should be focused on these effects on the changes in ET0 in the Songnen Grassland.
ET0 variations and its response to regional dry/wet conditions are of great importance for crop growing and natural vegetation [3,49]. For this purpose, the present study analyzed the long term variations of ET0, AI, and P for the growing season during 1960–2014. The results showed that ET0 and AI decreased as the P increased, especially during drought or wet years and vice versa (Figure 10). This was consistent with Madhu et al. [33], who reported that higher ET values were detected in the moderate and severe droughts years. It could be inferred from Equation (6) that AI has a positive correlation with ET0 and a negative correlation with P. The correlation was analyzed by the Pearson correlation analysis, and the results showed that AI has a significant positive correlation with ET0 (r = 0.766, p = 0.00), and it has a significant negative correlation with P (r = −0.977, p = 0.00). Figure 9 shows that all the trends of ET0, AI, and P were decreasing during the growing season from 1960–2014, and the decreasing rate of ET0 was higher than that of P. Thus, with the significant decreasing trend of ET0, the climate in the Songnen Grassland became slightly wetter from 1960–2014, and the trend will be continued if the decreasing trend of ET0 and no obvious change of precipitation persists. In short, under the decreasing ET0 conditions, agricultural and ecological water requirements will decrease, especially during the growing season, and agriculture irrigation water requirements will reduce as well. These could have a positive influence for vegetation growth, agricultural production, and ecology.

5. Conclusions

In this study, the spatial distributions of ET0, P, and their difference were obtained at different time scales. The temporal and spatial variations of ET0 were determined for 55 years of data from 21 stations in and around the Songnen Grassland, northeast China, during 1960–2014. The interannual variability of climatic variables was investigated during the study period, and a sensitivity analysis was conducted. The role of ET0 in regional dry/wet conditions was discussed based on the analysis of the relationship between ET0, P, and AI. The following conclusions can be drawn from this study.
(1)
The trend analysis of ET0 at different time scales shows an evident decreasing trend over the last 55 years, especially in the annual and spring periods. A break trend analysis shows that almost all considered climatological normal periods had experienced the decreasing trend, with a range of −2.415 to −0.003 mm per year. Abrupt changes were mainly detected in the early and mid-1990s in the annual, seasonal, and the growing season time series of ET0.
(2)
The spatial distributions of ET0 increased from the northeast to southwest in the annual, seasonal, and the growing season time series during 1960–2014. The spatial variations of ET0 indicated that the most significant decreasing trends were distributed in the eastern, northeastern, and central regions during the annual, spring, and growing season periods.
(3)
The interannual variability of climatic parameters indicated that the annual Max T, Ave T, and Min T displayed significant increasing trends at the 0.05 level (one-sided t-test), and significant decreasing trends were found for Ave RH, Win S, and Sun H. Ave RH was the dominant climate variable for the declining annual ET0 over the entire region, with the sensitivity decreasing from Max T, Win S, Sun H, Min T, to Ave T. Abrupt changes were detected in the annual time series of these variable; Ave RH in 1993, Max T in 1989, Win S in 1990, Sun H in 1987, Min T in 1983, and Ave T in 1987.
(4)
In general, the results of this study indicate that the regional drought/wetness condition became mildly wetter with decreasing ET0 during the growing season in the last 55 years. Regional climate drought has been alleviated in recent decades. These findings can serve as a reference for policy-makers for better planning and efficient use of agricultural water resources in the Songnen Grassland.

Acknowledgments

This study was financially supported by the National Key Technology R&D Program of China under Grant Nos. 2013BAK05B02 and 2013BAK05B01. The authors are grateful to the anonymous reviewers for their insightful and constructive comments for improving this manuscript.

Author Contributions

Jiquan Zhang was responsible for the original idea of the study; Caiyun Sun and Enliang Guo were responsible for the data compilation; Feng Zhang and Mengmeng Wang were responsible for the data processing and drawing; Qiyun Ma drafted the manuscript and all authors read and revised the final manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Stephens, G.L.; Hu, Y.X. Are climate-related changes to the character of global-mean precipitation predictable? Environ. Res. Lett. 2010, 5, 025209. [Google Scholar] [CrossRef]
  2. Escribano Frances, G.; Quevauviller, P.; San Martin Gonzalez, E.; Vargas Amelin, E. Climate change policy and water resources in the EU and Spain. A closer look into the Water Framework Directive. Environ. Sci. Policy 2017, 69, 1–12. [Google Scholar] [CrossRef]
  3. Croitoru, A.; Piticar, A.; Dragotă, C.S.; Burada, D.C. Recent changes in reference evapotranspiration in Romania. Glob. Planet. Chang. 2013, 111, 127–136. [Google Scholar] [CrossRef]
  4. Huo, Z.; Dai, X.; Feng, S.; Kang, S.; Huang, G. Effect of climate change on reference evapotranspiration and aridity index in arid region of China. J. Hydrol. 2013, 492, 24–34. [Google Scholar] [CrossRef]
  5. Fan, Z.; Thomas, A. Spatiotemporal variability of reference evapotranspiration and its contributing climatic factors in Yunnan Province, SW China, 1961–2004. Clim. Chang. 2013, 116, 309–325. [Google Scholar] [CrossRef]
  6. Lu, J.; Zhang, G.; Wu, F. Web-based Multi-Criteria Group Decision Support System with Linguistic Term Processing Function. IEEE Intell. Inform. Bull. 2005, 5, 35–43. [Google Scholar]
  7. Naderpour, M.; Lu, J.; Zhang, G. An intelligent situation awareness support system for safety-critical environments. Decis. Support Syst. 2014, 59, 325–340. [Google Scholar] [CrossRef]
  8. Shadmani, M.; Marofi, S.; Roknian, M. Trend Analysis in Reference Evapotranspiration Using Mann-Kendall and Spearman’s Rho Tests in Arid Regions of Iran. Water Resour. Manag. 2012, 26, 211–224. [Google Scholar] [CrossRef]
  9. Tabari, H.; Aghajanloo, M.B. Temporal pattern of aridity index in Iran with considering precipitation and evapotranspiration trends. Int. J. Clim. 2013, 33, 396–409. [Google Scholar] [CrossRef]
  10. Wang, Y.; Jiang, T.; Bothe, O.; Fraedrich, K. Changes of pan evaporation and reference evapotranspiration in the Yangtze River basin. Theor. Appl. Climatol. 2007, 90, 13–23. [Google Scholar] [CrossRef]
  11. Irmak, S.; Kabenge, I.; Skaggs, K.E.; Mutiibwa, D. Trend and magnitude of changes in climate variables and reference evapotranspiration over 116-yr period in the Platte River Basin, central Nebraska–USA. J. Hydrol. 2012, 420–421, 228–244. [Google Scholar] [CrossRef]
  12. Roderick, M.L.; Farquhar, G.D. Changes in New Zealand pan evaporation since the 1970s. Int. J. Climatol. 2005, 25, 2031–2039. [Google Scholar] [CrossRef]
  13. Bandyopadhyay, A.; Bhadra, A.; Raghuwanshi, N.S.; Singh, R. Temporal trends in estimates of reference evapotranspiration over India. J. Hydrol. Eng. 2009, 14, 508–515. [Google Scholar] [CrossRef]
  14. Liu, T.; Li, L.; Lai, J.; Liu, C.; Zhuang, W. Reference evapotranspiration change and its sensitivity to climate variables in southwest China. Theor. Appl. Climatol. 2016, 125, 1–10. [Google Scholar] [CrossRef]
  15. Dinpashoh, Y.; Jhajharia, D.; Fakheri-Fard, A.; Singh, V.P.; Kahya, E. Trends in reference crop evapotranspiration over Iran. J. Hydrol. 2011, 375, 65–77. [Google Scholar] [CrossRef]
  16. Liu, Y.; Zhuang, Q.; Pan, Z.; Miralles, D.; Tchebakova, N.; Kicklighter, D.; Chen, J.; Sirin, A.; He, Y.; Zhou, G.; et al. Response of evapotranspiration and water availability to the changing climate in Northern Eurasia. Clim. Chang. 2014, 126, 413–427. [Google Scholar] [CrossRef]
  17. Zheng, S.; Qin, Z.; Zhang, W. Drought variation in Songnen Plain and its response to climate change. Chin. J. Agrometeorol. 2015, 36, 640–649. [Google Scholar]
  18. Zhang, Y.; Deng, J.; Guan, D.; Jin, C.; Wang, A.; Wu, J.; Yuan, F. Spatiotemporal changes of potential evapotranspiration in Songnen Plain of Northeast China. Chin. J. Appl. Ecol. 2011, 22, 1702–1710. [Google Scholar]
  19. Limin, S.G.; Oue, H.; Takase, K. Estimation of areal average rainfall in the mountainous Kamo River Watershed, Japan. J. Agric. Meteorol. 2015, 71, 90–97. [Google Scholar] [CrossRef]
  20. Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop Evapotranspiration-Guidelines for Computing Crop Water Requirements—FAO Irrigation and Drainage; FAO: Rome, Italy, 1998; p. 56. [Google Scholar]
  21. Mosaedi, A.; Sough, M.G.; Sadeghi, S.; Mooshakhian, Y.; Bannayan, M. Sensitivity analysis of monthly reference crop evapotranspiration trends in Iran: A qualitative approach. Theor. Appl. Climatol. 2017, 128, 857–873. [Google Scholar] [CrossRef]
  22. Liu, C.; Zhang, D.; Liu, X.; Zhao, C. Spatial and temporal change in the potential evapotranspiration sensitivity to meteorological factors in China (1960–2007). J. Geogr. Sci. 2012, 22, 3–14. [Google Scholar] [CrossRef]
  23. Kendall, M.G. Rank Correlation Methods; Griffin: London, UK, 1948. [Google Scholar]
  24. Mann, H.B. Nonparametric tests against trend. Econometrica 1945, 13, 245–259. [Google Scholar] [CrossRef]
  25. Wang, W.; Zhu, Y.; Xu, R.; Liu, J. Drought severity change in China during 1961–2012 indicated by SPI and SPEI. Nat. Hazards 2015, 75, 2437–2451. [Google Scholar] [CrossRef]
  26. Zhang, Q.; Zhang, J. Drought hazard assessment in typical corn cultivated areas of China at present and potential climate change. Nat. Hazards 2016, 81, 1323–1331. [Google Scholar] [CrossRef]
  27. Yan, T.; Shen, Z.; Bai, J. Spatial and temporal changes in temperature, precipitation, and streamflow in the Miyun Reservoir Basin of China. Water 2017, 9, 78. [Google Scholar] [CrossRef]
  28. Tabari, H.; Taye, M.T.; Willems, P. Statistical assessment of precipitation trends in the upper Blue Nile River basin. Stoch. Environ. Res. Risk Assess. 2015, 29, 1751–1961. [Google Scholar] [CrossRef]
  29. Palizdan, N.; Falamarzi, Y.; Huang, Y.F.; Lee, T.S. Precipitation trend analysis using discrete wavelet transform at the Langat River Basin, Selangor, Malaysia. Stoch. Environ. Res. Risk Assess. 2017, 31, 853–877. [Google Scholar] [CrossRef]
  30. Liang, L.; Li, L.; Liu, Q. Temporal variation of reference evapotranspiration during 1961–2005 in the Taoer River basin of Northeast China. Agric. For. Meteorol. 2010, 150, 298–306. [Google Scholar] [CrossRef]
  31. Sen, P.K. Estimates of the regression coefficient based on Kendall’s tau. J. Am. Stat. Assoc. 1968, 63, 1379–1389. [Google Scholar] [CrossRef]
  32. Salmi, T.; Maatta, A.; Anttila, P.; Ruoho-Airola, T.; Amnell, T. Detecting Trends of Annual Values of Atmospheric Pollutants by the Mann-Kendall Test and Sen’s Slope Estimates—The Excel Template Application MAKESENS; Publications on Air Quality 31; Report Code FMIAQ-31; Universitas Gadjah Mada: Yogyakarta, The Republic of Indonesia, 2002. [Google Scholar]
  33. Madhu, S.; Kumar, T.V.L.; Barbosa, H.; Rao, K.K.; Bhaskar, V.V. Trend analysis of evapotranspiration and its response to droughts over India. Theor. Appl. Climatol. 2015, 121, 41–51. [Google Scholar] [CrossRef]
  34. Arguez, A.; Vose, R.S. The Definition of the Standard WMO Climate Normal: The Key to Deriving Alternative Climate Normals. Bull. Am. Meteorol. Soc. 2011, 92, 699–704. [Google Scholar] [CrossRef]
  35. Zhao, L.; Xia, J.; Sobkowiak, L.; Li, Z. Climatic characteristics of reference evapotranspiration in the Hai River Basin and their attribution. Water 2014, 6, 1482–1499. [Google Scholar] [CrossRef]
  36. Zheng, C.; Wang, Q. Spatiotemporal pattern of the global sensitivity of the reference evapotranspiration to climatic variables in recent five decades over China. Stoch. Environ. Res. Risk. Assess. 2015, 29, 1937–1947. [Google Scholar] [CrossRef]
  37. Zhan, X.; Kustas, W.P.; Humes, K.S. An intercomparison study on models of sensible heat flux over partial canopy surfaces with remotely sensed surface temperature. Remote Sens. Environ. 1996, 58, 242–256. [Google Scholar] [CrossRef]
  38. Su, X.; Singh, V.P.; Niu, J.; Hao, L. Spatiotemporal trends of aridity index in Shiyang River basin of northwest China. Stoch. Environ. Res. Risk. Assess. 2015, 29, 1571–1582. [Google Scholar] [CrossRef]
  39. Thornthwaite, C.W. An approach toward a rational classification of climate. Geogr. Rev. 1948, 38, 55–89. [Google Scholar] [CrossRef]
  40. Zhang, K.; Pan, S.; Zhang, W.; Xu, Y.; Cao, L.; Hao, Y.; Wang, Y. Influence of climate change on reference evapotranspiration and aridity index and their temporal-spatial variations in the Yellow River Basin, China, from 1961 to 2012. Quat. Int. 2015, 380–381, 75–82. [Google Scholar] [CrossRef]
  41. Ashraf, M.; Routray, J.K. Spatio-temporal characteristics of precipitation and drought in Balochistan Province, Pakistan. Nat. Hazards 2015, 77, 229–254. [Google Scholar] [CrossRef]
  42. Li, X.; Gemmer, M.; Zhai, J.; Liu, X.; Su, B.; Wang, Y. Spatio-temporal variation of actual evapotranspiration in the Haihe River Basin of the past 50 years. Quat. Int. 2013, 304, 133–141. [Google Scholar] [CrossRef]
  43. Piticar, A.; Mihăilă, D.; Lazurca, L.G.; Bistricean, P.; Puţuntică, A.; Briciu, A. Spatiotemporal distribution of reference evapotranspiration in the Republic of Moldova. Theor. Appl. Climatol. 2016, 124, 1133–1144. [Google Scholar] [CrossRef]
  44. Yin, Y.; Wu, S.; Chen, G.; Dai, E. Attribution analyses of potential evapotranspiration changes in China since the 1960s. Theor. Appl. Climatol. 2010, 101, 19–28. [Google Scholar] [CrossRef]
  45. Dong, B.; Sutton, R.T.; Chen, W.; Liu, X.; Lu, R.; Sun, Y. Abrupt summer warming and changes in temperature extremes over Northeast Asia since the mid-1990s: Drivers and physical processes. Adv. Atmos. Sci. 2016, 33, 1005–1023. [Google Scholar] [CrossRef]
  46. Liu, S.; Yang, S.; Lian, Y.; Zheng, D.; Wen, M.; Tu, G.; Shen, B.; Gao, Z.; Wang, D. Time–frequency characteristics of regional climate over northeast China and their relationships with atmospheric circulation patterns. J. Clim. 2010, 23, 4956–4972. [Google Scholar] [CrossRef]
  47. Gong, D.Y.; Wang, S.W. Influence of Arctic Oscillation on Winter Climate over China. J. Geogr. Sci. 2003, 13, 208–216. [Google Scholar]
  48. Li, J.; Jiao, M.; Hu, C.; Li, F.; Zhang, X.; Zhang, Q.; Wang, Y.; Zhu, X. Characteristics of summer temperature and its impact factors in Northeast China from 1951 to 2012. J. Meteorol. Environ. 2016, 32, 74–83. [Google Scholar]
  49. Espadafor, M.; Lorite, I.J.; Gavilan, P.; Berengena, J. An analysis of the tendency of reference evapotranspiration estimates and other climate variables during the last 45 years in Southern Spain. Agric. Water Manag. 2011, 98, 1045–1061. [Google Scholar] [CrossRef]
Figure 1. Location of the Songnen Grassland in China and the meteorological stations considered.
Figure 1. Location of the Songnen Grassland in China and the meteorological stations considered.
Water 09 00316 g001
Figure 2. Spatial distribution of ET0, P, and the difference between the two for the period from 1960–2014.
Figure 2. Spatial distribution of ET0, P, and the difference between the two for the period from 1960–2014.
Water 09 00316 g002
Figure 3. Linear trend and break trend analysis (for every 30 years) of ET0 from 1960 to 2014. The black dashed line is the average value of the four climatological normal periods. Water 09 00316 i001 indicates the detected change point.
Figure 3. Linear trend and break trend analysis (for every 30 years) of ET0 from 1960 to 2014. The black dashed line is the average value of the four climatological normal periods. Water 09 00316 i001 indicates the detected change point.
Water 09 00316 g003aWater 09 00316 g003b
Figure 4. Frequency of ET0 trends for the 21 stations based on the Mann-Kendall (MK) test in the Songnen Grassland during 1960–2014 (%).
Figure 4. Frequency of ET0 trends for the 21 stations based on the Mann-Kendall (MK) test in the Songnen Grassland during 1960–2014 (%).
Water 09 00316 g004
Figure 5. Spatial distribution of trends and their magnitudes in ET0 over the period from 1960–2014. The Sig. indicates a significant trend at the 0.05 level. The black line is the contour line of the magnitude of the ET0 trend.
Figure 5. Spatial distribution of trends and their magnitudes in ET0 over the period from 1960–2014. The Sig. indicates a significant trend at the 0.05 level. The black line is the contour line of the magnitude of the ET0 trend.
Water 09 00316 g005aWater 09 00316 g005b
Figure 6. Linear trend and break trend analysis (for every 30 years) of the climatic parameters from 1960 to 2014. Water 09 00316 i001 indicates the detected change point.
Figure 6. Linear trend and break trend analysis (for every 30 years) of the climatic parameters from 1960 to 2014. Water 09 00316 i001 indicates the detected change point.
Water 09 00316 g006
Figure 7. Frequency of climatic parameter trends in the Songnen Grassland over the period from 1960–2014 (%).
Figure 7. Frequency of climatic parameter trends in the Songnen Grassland over the period from 1960–2014 (%).
Water 09 00316 g007
Figure 8. Spatial distribution of the average annual sensitivities of the climatic variables and the result of the sensitivity comparison in 13 inside stations during 1960–2014.
Figure 8. Spatial distribution of the average annual sensitivities of the climatic variables and the result of the sensitivity comparison in 13 inside stations during 1960–2014.
Water 09 00316 g008
Figure 9. Comparison of the temporal trend of ET0, AI, and P during 1960–2014. D, the drought year; W, the wet year. The slopes are examined by linear regression.
Figure 9. Comparison of the temporal trend of ET0, AI, and P during 1960–2014. D, the drought year; W, the wet year. The slopes are examined by linear regression.
Water 09 00316 g009
Figure 10. P-P plot of the regression residual standard error for ET0, AI, and P.
Figure 10. P-P plot of the regression residual standard error for ET0, AI, and P.
Water 09 00316 g010
Table 1. Summary of the sensitive climatic variables that contributed to the changes in ET0 at the stations during 1960–2014.
Table 1. Summary of the sensitive climatic variables that contributed to the changes in ET0 at the stations during 1960–2014.
StationOrder and Trend of the Sensitivity for the Climatic Variables in ET0
KeshanAve T−5.55 *Ave RH−3.54 *Max T−3.59 *Win S−0.81Sun H3.85 *Min T0.94
FuyuAve RH−2.38 *Max T−3.17 *Sun H2.06 *Min T1.56Win S0.07Ave T1.10
QiqihaerAve RH−4.56 *Max T−4.15 *Win S1.31Sun H2.06 *Min T4.24 *Ave T0.00
MingshuiMin T−6.90 *Ave RH−3.96 *Max T−4.85 *Win S0.64Sun H1.51Ave T0.00
TailaiAve RH−3.89 *Max T−3.12 *Win S4.02 *Sun H1.29Min T3.66 *Ave T0.00
AndaAve RH2.83 *Max T−3.56 *Win S0.20Sun H4.07 *Min T0.40Ave T0.00
BaichengAve RH−3.05 *Max T−3.03 *Win S1.19Sun H1.80Min T2.08 *Ave T0.00
Qian’anAve RH−2.90 *Max T−2.87 *Win S2.53 *Sun H0.28Min T1.68Ave T0.00
Qian GorlosAve RH−6.27 *Max T−4.66 *Win S0.96Sun H3.53*Ave T−4.72 *Min T2.31 *
TongyuAve RH−2.95 *Max T−1.74Win S3.03 *Sun H−0.15Min T1.06Ave T0.00
ChanglingAve RH−3.27 *Max T−2.63 *Win S4.01 *Sun H−1.26Min T2.42 *Ave T0.00
Fuyu CityAve RH−5.62 *Max T−4.37 *Win S2.42 *Sun H1.28Min T3.14 *Ave T0.00
ChangchunMax T3.11 *Ave RH−6.11 *Win S1.97 *Sun H3.47 *Min T2.24 *Ave T0.00
RegionAve RH−4.07 *Max T−4.18 *Win S2.90 *Sun H3.92 *Ave T−6.94 *Min T−7.93 *
Notes: The order means the ranking of the sensitive climatic variables (descending) at each station. The trends of the sensitivity for the climatic variables were detected by the MK test. * Indicates a significance level of 0.05.
Water EISSN 2073-4441 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top