1. Introduction
Water surface temperature is one of the basic physical properties of water (Wrzesiński et al. 2015), and its changes have important influences on freshwater ecosystems, such as changes in species distribution and convenience of species invasion (Ptak et al. 2019). With global warming, open water has received much attention due to its high sensitivity to climate change (Qin et al. 2010). Among the ways in which climate change information is transmitted to open water, the most noticeable impact on open water is reflected in water surface temperature (Yan and Zheng 2015). Recent studies have focused on the relationship between ocean surface temperature and global climate change (Chen et al. 2018; Carrillo et al. 2018; Yan et al. 2016; Shirvani et al. 2015), while there are relatively few studies on inland water surface temperature (IWST). According to a National Aeronautics and Space Administration (NASA) and a National Science Foundation–funded study, climate change is rapidly warming lakes around the world (NASA 2015). For a long time, the research on the variation characteristics of water body temperature in China focuses on the analysis of single reservoir, lake or small area. Moreover, the temporal resolution of the data used is usually monthly, and the spatial resolution is usually more than 4 km, which cannot reveal the variation characteristics of water surface temperature on a macroscale. Traditional water surface temperature data are mainly collected from in situ measurements. For example, Ptak et al. (2019) collected daily measurements of water temperature and air temperature in the Warta River using data from six hydrological stations, and analyzed the direction of water temperature fluctuation in the Warta River from 1960 to 2009. However, in situ measurement data are usually very sparse in space, and few in situ measurements are available at suitable locations and times (Trumpickas et al. 2009). Recently, satellite based thermal data, such as Scanning Multichannel Microwave Radiometer (SMMR) and Special Sensor Microwave Imager (SSM/I), are increasingly used to monitor the surface temperature of inland water bodies. They are usually Advanced Very High Resolution Radiometer (AVHRR), Along Track Scanning Radiometer (ATSR), Landsat, MODIS, and infrared photography. MODIS temperature data are the best among the current remote sensing data of temperature data, with high spatial and temporal resolution, large spectral range, and high radiation sensitivity, and the thermal infrared channel setting is reasonable. In contrast, the resolution of NOAA and other satellites is coarser and unavailable. MODIS products can be used to measure inland water surface temperature with considerable accuracy.
Due to the different variation characteristics and responses of individual IWST to climate change across China, it was not clear how China’s inland waters have responded to climate change in the past, and there is no assessment of the temporal and spatial variation characteristics of water surface temperature using a consistent method on a national scale in China. Here, we adopt a unified method (the wavelet transform) to study China’s IWST changes using satellite remote sensing data (i.e., MODIS) and near-surface air temperature data from 2000 to 2015. Compared with other studies, this study has the advantages of large research scope, high spatial–temporal resolution of temperature data and novel analysis methods. This study will help understand China’s inland wetland responses to global climate change and provide scientific support for the management of wetland ecosystems.
2. Study area and data sources
a. Study area
The distribution of China’s inland water came from Moderate Resolution Imaging Spectroradiometer Land Water Mask (MOD44W) data products. The data product has a spatial resolution of 250 m, a time resolution of 1 year, and a time range of 2000–15. To keep consistent with the spatial resolution of temperature data, the nearest neighbor resampling method was used to convert the data into 1-km spatial resolution. Considering the water dynamics, we overlapped the yearly water maps of MOD44W data products to obtain a stable water distribution map. In a 3 × 3 grid, the middle patch is considered to be stable. Therefore, considering the stability of the water body, we selected the water patches with an area of more than 9 km2 that lasted for 16 years from 2000 to 2015 as our study area (Fig. 1). There was a total of 733 stable water patches with an area of more than 9 km2 in China.
Distribution of lakes in the Great Lakes region of China (China’s lakes are grouped into five geographical distribution areas: Northeast Lake region, East Lake region, Yungui Lake region, Qinghai Lake region, and Mengxin Lake region (Nanjing Institute of Geography and Limnology 2015).
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
As a representative of inland water, a lake is the junction of the interactions between the various elements of Earth’s surface ecosystem and has the function of recording environmental changes (Moser et al. 2019). The differentiation of the regional natural environment makes the natural characteristics of lakes in China show noticeable regional differences. According to the lake area and the geographical distribution area, we selected 17 representative lakes (Jingpo Lake, Chagan Lake, Xingkai Lake, Dianshan Lake, Weishan Lake, Chaohu Lake, Taihu Lake, Lugu Lake, Fuxian Lake, Dongcuo Lake, Lake Manasarovar, Zaling Lake, Qinghai Lake, Kanas Lake, Sailimu Lake, Aibi Lake, and Hulun Lake) across those five regions to study the temporal variation of water surface temperature.
b. Data sources
The water surface temperature came from MODIS surface temperature product MOD11A1 data (Wan et al. 2015) with a spatial resolution of 1 km and time resolution of 1 day. The water mask data came from global surface water product MOD44W version 6 with a spatial resolution of 250 m, which is available annually from 2000 to 2015. We selected the daytime data of MOD11A1 land surface temperature products from 2000 to 2015. We used the stable water map, which is derived from MOD44W, to mask the surface temperature product MOD11A1 to obtain the daily surface temperature of China’s inland water.
The water surface temperature of one water body is usually not identical. In this study, the mode value of the water temperature grid value was selected to represent the temperature of the whole water patch. Due to the limitation of data sources, some of the derived water surface temperatures were null. Therefore, the water surface temperature was interpolated in time series based on the date to obtain the daily water surface temperature from 2000 to 2015. As a result, we acquired 733 water surface temperature records during the period and then used triangulated irregular networks (TIN) space interpolation to produce spatially continuous water surface temperature distribution information for the past 16 years. TIN interpolation is a common tool in GIS. A common TIN algorithm is called Delaunay triangulation. It attempts to create a surface composed of the triangles of the nearest adjacent points. To do this, a circle around the selected sample points is created and its intersection is connected to a nonoverlapping and as compact as possible triangular network. TIN method is more reasonable than inverse distance weighted interpolation.
The air temperature data were based on the near-surface temperature elements of China’s regional ground meteorological elements dataset (Yang and He 2016). It is a set of near-surface meteorological and environmental element reanalysis datasets with a temporal resolution of 3 h and a spatial resolution of 0.1°. It was based on the internationally existing Princeton reanalysis data, Global Land Data Assimilation System (GLDAS) data, GEWEX-SRB radiation data, and Tropical Rainfall Measuring Mission (TRMM) satellite precipitation data as a background field, and was made up of conventional meteorological observation data from the China Meteorological Administration.
3. Methods
a. Wavelet transform
Among those methods used for time series analysis, serial-correlation analysis and Fourier transform are the most traditional and typical. When applied to hydrologic time series which operate over multitemporal scales and show nonstationary characteristics, these methods usually cannot meet practical needs due to their limited applicable conditions (Sang 2013). Comparatively, wavelet analysis can reveal the local characteristics of time series in time domain and frequency domain at the same time, which is a more powerful method of time series analysis, and more suitable to study hydrological time series (Sang et al. 2013). The wavelet transform is a signal analysis method (Li 2016) that can automatically adapt to the requirements of time–frequency analysis (Zhang et al. 2013). It can decompose the time series into a high-frequency part and a low-frequency part. The high-frequency part can be used to extract the random and periodic items of the time series, and the remaining low-frequency part can be used to extract the trend items of the time series (Zhang 2018). It has been widely used in time series feature extraction (Praveen and Efi 1997; Liu, 2012; Sang et al. 2013; Kędra and Wiejaczka 2018; Zhang 2018). Liu (2012) used wavelet transform to extract the trend term of hydrological time series, and compared the wavelet transform with the traditional methods (trend regression method, moving average method, and Mann–Kendall test method) to extract the trend items of time series. The results showed that the theory of extracting trend term of time series based on wavelet transform was better.
When applying the wavelet transform to extract features of time series, the wavelet basis function and decomposition level should be determined. The wavelet basis function is related to whether it has regularity, orthogonality, symmetry, vanishing moment, and tight support. The decomposition level is determined based on the complexity of the signal. The more complex the signal is, the more decomposition layers should be selected (Zhang et al. 2014).
In this paper, the wavelet transform was mainly used to extract trend terms of the time series. Due to noise pollution, useful trend information cannot be directly seen in time domain. To extract the trend term effectively, the db8 wavelet with a higher vanishing moment and lower compactness support is needed to make the time series smooth, which is conducive to trend judgment (Zhang et al. 2014). The time series of the study was the daily lake temperature in 16 years, so the signal was complex. In this study, the decomposition level was selected as 11 layers. The low-frequency information a11 was used to obtain the interannual variation trend of lake temperature and air temperature above the lake. When the study time series was the midpoint of the 0°C isotherm every year for 16 years, the signal was simple and the decomposition level selected at this time was three layers. The low-frequency information a3 is used to obtain the trend information of interannual variation of the midpoint’s latitude and longitude of the 0°C isotherm in China.
b. Ordinary least squares linear regression
If SLOPE is positive, the freezing information shows an increasing trend in n years; otherwise, it shows a decreasing trend. If SLOPE is equal to zero, there is no significant change in freezing information during the study period. In addition, the greater the absolute value of SLOPE is, the more obvious the changing trend.
c. Development of the 0°C isotherm
We used the 0°C isotherm to study the spatial variation of IWST based on daily water surface temperature with raster format. We first produced the water surface temperature isotherm by using the GIS tool (ArcGIS 10.2) and then selected the 0°C isotherm to analyze the change of IWST. Although several 0°C isotherms existed across China for one year, the longest 0°C isotherm in the northeast–southwest direction always existed. Therefore, the longest 0°C isotherm was selected for analysis. To study the spatial changes of IWST, we chose the midpoint and length of the 0°C isotherm to characterize the spatial distribution and change trend of the 0°C isotherm.
4. Results and analysis
a. Spatial variation characteristics of IWST
1) Seasonal variation
Taking the year of 2015 as an example, we studied the seasonal variation of IWST through spatial distribution of the 0°C isotherm for a year. The schematic diagram of the 0°C isothermal line of spring (March–May), summer (June–August), autumn (September–November), and winter (December–February) in Chinese mainland is shown in Fig. 2. The dates in this study were calculated according to Julian Day (1 January is the first day, 2 January is the second day, and 31 December is the 365th day). Then, the wavelet transform was applied to the midpoint position information of the longest 0°C isotherm. The results were as follows: from the first day (1 January), the midpoint of the 0°C isotherm moved from about 35°N to about 50°N on the 118th day (28 April), and from the 300th day (27 October), it gradually moved back to 35°N on the 365th day (31 December) (Fig. 3). At the same time, the midpoint of the 0°C isotherm moved from about 105° to 130°E (the 98th day, 8 April), and then gradually moved back westward to about 105°E. It can be seen that the 0°C isotherm moved first to the northeast direction, and then gradually moved back, which showed a cycle movement of the 0°C isotherm in a year, characterizing the seasonal variation of China’s IWST (Fig. 3). It also showed a fact that water bodies in the northern and eastern regions are frozen earlier and melting later. Because the surface temperature of all of the water bodies in China was higher than 0°C from the 120th to 300th day for 2015, no 0°C isotherm existed during this period (Fig. 3).
Diagram of the 0°C isotherm of Chinese water bodies in four seasons.
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
(a) North latitude, (b) east longitude, and (c) length variation of the 0°C isotherm of China’s inland water for 2015.
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
Similarly, the length of the longest 0°C isotherm first remained relatively stable from the first day to the 75th day, and then it began shortening sharply until disappeared on the 105th day. From the 300th day, it appeared and grew rapidly until the 320th day. Then it became stable until the end of the year (Fig. 3).
2) Interannual variation
Taking the day of the winter solstice (the 356th day) as the representative day, we studied the spatial changes of China’s IWST from 2000 to 2015 using the location of midpoint of the longest 0°C isotherm. As can be seen from Fig. 4, it gradually shifted northward by approximately 0.09° in the past 16 years. In the longitude direction, it first moved eastward by about 2° before 2008, and then gradually moved westward by about 1°. Therefore, it showed first northeastern moving trend and then northwestern moving trend during this period, implying the impact of climate warming on China inland water bodies. The midpoint’s altitude of the 0°C isotherm first showed decreasing before 2008, and then increasing slightly between 2008 and 2015 (Fig. 4).
Changes in (a) latitude of the midpoint, (b) longitude of the midpoint, (c) length, and (d) altitude of the midpoint of the 0°C isotherm of China’s water body during the 2000–15 winter solstice.
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
During the past 16 years, the length of the 0°C isotherm showed a downward trend. By 2008, the length was reduced to a minimum and then increased slightly (Fig. 4). This indicated that the range of 0°C water body tended to decrease before 2008 and gradually grew after 2008, which was related to the spatial distribution of China’s inland waters. The turning of the above changes could imply the altering of the warming trend in the year of 2018.
b. Temporal variation characteristics
Changes of the water freezing date and melting date can directly reflect impacts of climate changes on inland waters. Regarding the date of freezing and complete melting, the method in reference (Yu et al. 2018) was adopted. The indicator of ice froze phenomenon was that temperature of the water patches was lower than 0°C for the first time in a year. The first day of ice frozen phenomenon over 5 days for the first time in a year was the water freezing date. And the last day of ice froze phenomenon over 5 days for the last time before 30 June in the next year was the ice melting date.
The trend of the start date of water freezing was clearly delayed during 2001–15 with an average rate of −1.5 days yr−1 (Fig. 5). Yet, the trend of the date of water complete melting was not obvious. From 2000 to 2015, both the first occurrence of icing and the final appearance of ice melting were located on the Tibetan Plateau (Fig. 5). Among them, there were 11 years that the water bodies with the earliest freezing phenomenon were located in the southern part of the Tibetan Plateau, while water bodies with the final ice-melting phenomenon were located in the northern part of the Tibetan Plateau in the whole 16 years.
Location and interannual variation of the date of (left) water freezing and (right) complete melting from 2000 to 2015.
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
From 2000 to 2015, the water surface temperature of the 17 representative lakes across China showed the same variation trend which rose first and then fell, but the time that the inflection point occurred and the change range of water surface temperature after the inflection point was different (Fig. 6). The water surface temperature of most (16) lakes with various different sizes experienced an inflection point in the year 2010. By contrast, the year of the inflection point in the water surface temperature of Lake Manasarovar was the year 2008. In general, after the inflection point of interannual variation, lake temperature in Mengxin Lake region decreased the most, followed by Northeast Lake region, and lake temperature in Yungui Lake region decreased the least.
Interannual variation of lake surface temperature and air temperature in the Great Lakes region of China (the curves are change trend information obtained by wavelet transform).
Citation: Journal of Hydrometeorology 22, 2; 10.1175/JHM-D-20-0104.1
We compared the interannual changes of water surface temperature of various lakes and air temperatures to explore the different response characteristics of lake surface temperature to air temperature across China. In general, the lake surface temperature and air temperature across China showed the same variation trend in which rises first and then falls. However, the lake surface temperature and air temperature had synchronous changes in those lakes, including Xingkai Lake, Dianshan Lake, Weishan Lake, Taihu Lake, Lugu Lake, Dongcuo Lake, Qinghai Lake, Aibi Lake, and Hulun Lake. Specifically, the inflection point of both the lake surface temperature and air temperature change of the above lakes occurred in the year 2010. In contrast, the inflection points of interannual changes in lake surface temperature and air temperature in the other eight lakes were inconsistent. The inflection point of air temperature change appeared earlier than the lake surface temperature change in Jingpo Lake, Chagan Lake, Chaohu Lake, Lake Manasarovar, Zaling Lake, Kanas Lake, and Sailimu Lake, while Fuxian Lake showed the opposite phenomenon.
5. Discussion
The intra-annual movement path of the 0°C isotherm of IWST, which first moved northeast and then gradually moved back, reflected the impact of seasonal climate on IWST across China. In the meantime, the spatial distribution of 0°C isotherm of water was also consistent with geomorphic distribution in China. However, the interannual variation change showed a trend of eastward movement, which indicated that the IWST in China was more influenced by the geomorphic pattern than zonal impact. Ding and Elmore (2015) found that water surface and air temperatures increased from 1984 to 2007 at all temperature stations analyzed, but the strength of this increase varied by location. Water bodies in China that first froze and finally melted were both located in the Tibetan plateau, which was consistent with the characteristics of the Tibetan Plateau’s cold climate and long freeze period of lakes (Nanjing Institute of Geography and Limnology 2015). In accordance with Trumpickas et al. (2009), surface water temperatures typically exhibit linear warming through the spring, form a plateau in midsummer and then exhibit linear cooling in fall. Lake-specific warming and cooling rates vary little from year to year while plateau values vary substantially across years.
The delayed date of water freezing reflected the more effect of climate warming in winter than that in spring. Yu et al. (2011) analyzed the annual and seasonal variation characteristics of the national average temperature from 1951 to 2009 and showed that the trend of winter average temperature rise is the most obvious in China, followed by spring. Xu et al. (2018) also found that the winter temperature in northeast China is on the rise.
There was a horizontal tonality of water surface temperature in China, which decreased from south to north (Arai and Pu 2009). Therefore, the interannual movement of the 0°C isotherm of water on Winter Solstice has shown a northward moving trend in the past 16 years, which further validated the response of inland water to climate warming. From 2000 to 2015, the altitude of the 0°C isotherm of China’s water body showed a continuous decreasing downward trend and the reduced altitude is about 39 m. This was the possible reason for the eastward shift (about 1°) of the 0°C isotherm of water in China over the past 16 years.
The water surface temperature of four lakes in the Qinghai Lake region in 2000–15 showed an overall variation trend of increasing first and then decreasing. Among them, the water surface temperature of Qinghai Lake showed an upward trend in 2000–10, which was consistent with the results of Xiao et al. (2013), who reported that the water surface temperature of Qinghai Lake was increasing in 2001–10 based on MODIS data and linear regression method. The inflection point of water surface temperature in Lake Manasarovar was 2008, that was, the lake temperature showed an upward trend before 2008 and then it showed a downward trend. This may be because the air temperature in the basin of Lake Manasarovar increased from 1975 to 2008 and the precipitation decreased, which made the glaciers melted more (La et al. 2012). More snow and ice melted into the lake, resulting in a downward trend in water surface temperature after 2008.
The inconsistency between the results of this study and those of previous studies indicated that differences in research data may lead to differences in results. Wan et al. (2018) used MODIS data to study the changes in the nocturnal surface temperature of lakes in the Tibetan Plateau from 2001 to 2015. The results showed that 70% of lakes have an upward trend in the past 15 years. Wan et al. (2018) obtained the interannual variation trend of lake temperature based on the yearly average value of lake temperature, while our research employed the daily lake temperature to obtain the overall variation trend of lake temperature from 2000 to 2015. Therefore, the difference in the temporal resolution of the dataset could lead to the different results.
According to the result of this study, the temperature of lakes in China has shown an upward trend from 2000 to 2010. This was consistent with trends in the temperature of inland water bodies in other countries around the world. Schneider et al. (2009) used satellite observations and in situ observations to study the change trend of lake temperature in inland North America. They found that lakes in California and Nevada had a rapid warming trend from 1991 to 2008. By summarizing the previous research results, Dokulil (2014) also found that most inland waters in Europe showed a warming trend from 1960s to 2010.
The decreasing range of lake surface temperature after the inflection point was related to the geographical location of the lakes. For example, the lake temperature in the Mengxin Lake region and Northeast Lake region located in northern China declined much more than that of lakes in the Yungui Lake region, which was located in the south of China. By comparing the interannual variation of water surface temperature and air temperature (Fig. 6), we can find that the inflection point of lake surface temperature in Fuxian Lake occurred earlier than that of air temperature. The inflection year of Fuxian Lake surface temperature change was 2009, while its air temperature change inflection was in the year 2010. After that, although the water surface temperature showed a downward trend, the decline was small, and the water surface temperature gradually stabilized. This may be because Fuxian Lake is a deep-water lake, and its ability to regulate temperature is much greater than that of shallow-water lake.
Past studies have argued that the air temperature of Aibi Lake between 2000 and 2014, the air temperature of Hulu Lake between 1960 and 2014, and the air temperature of Taihu Lake between 1958 and 2017 were all increasing (Zhao et al. 2017; Wu et al. 2016, 2019). The air temperature data used in those studies were derived from a limited number of ground stations around the lakes, which was not representative and had limitations in the case of uneven surfaces. By contrast, the overall change trend of air temperature above those lakes, which was derived from reanalysis data by interpolating the meteorological station observation data of China Meteorological Administration and existing Princeton reanalysis data, was first increasing and then decreasing between 2000 and 2015. The downward trend after the inflection year could be caused by the increased evaporation rate above lakes, which was the result of global warming. As products of global warming, El Niño and La Niña had impacts on climate change in China (Glantz 2000). During the occurrence of La Niña, the strengthening of the global pressure field made the northerly airflow prevailed in China’s winter. Therefore, cold wave and strong wind occurred frequently, and the ice situation in the Yellow Sea and Bohai Sea was generally heavy (Zhang and Meng 2005). The transition year of El Niño and La Niña was 2010–11 (Zahidul 2018). The occurrence of the La Niña phenomenon could also be the possible cause of the decreasing trend of both lake temperature and air temperature above lakes in China after 2010.
6. Conclusions
Water surface temperature is a direct indication of climate change. To understand the responses of China’s inland waters to climate change on a national scale, we employed the daily temperature products from satellite data between 2000 and 2015 to study the spatial and temporal characteristics of China’s inland surface water temperature using the wavelet transform method. Some conclusions are as follows:
The change of the 0°C isotherm of surface water in China had obvious seasonal variation. Both the midpoint location and length of the longest 0°C isotherm can characterize the seasonal changes of surface temperature for China’s inland waters.
From 2000 to 2015, the freezing period of inland water in China was obviously postponed, and the first and last ice melting occurred in the Tibetan Plateau.
The seasonal climate and geomorphic pattern in China have an impact on the surface temperature of inland water body. The interannual variation of the 0°C isotherm on the winter solstice shows a northward trend, which further verifies the response of inland water to climate warming.
The change of water surface temperature in China and its response to global climate change have enhanced the understanding of the response of wetlands to climate change. The degree of response is spatially and temporally distinct and closely related to the physical and geographic properties of the water.
Acknowledgments
This study was supported by the National Key Research and Development Program of China (2017YFA0603004), the National Natural Science Foundation of China (41971390). MODIS data products (MOD44W, MOD11A1) from 2000 to 2015 were obtained from https://lpdaac.usgs.gov, maintained by the NASA Earth Observing System Data and Information System (EOSDIS) Land Processes Distributed Active Archive Center (LP DAAC) at the USGS/Earth Resources Observation and Science (EROS) Center, Sioux Falls, South Dakota. Climate data (China’s regional ground meteorological elements dataset) were developed by Cold and Arid Regions Science Data Center at Lanzhou (https://data.tpdc.ac.cn/en/data/7a35329c-c53f-4267-aa07-e0037d913a21, doi:10.3972/westdc.002.2014.db). The authors declare that they have no conflict of interest.
REFERENCES
Arai, T., and P. Pu, 2009: A preliminary study on the water temperature and freezing of Lake Suwa in Japan and shallow lakes in eastern China. Japan J. Limnol., 48, 225–230, https://doi.org/10.3739/rikusui.48.225.
Carrillo, C. M., C. L. Castro, G. Garfin, H. I. Chang, M. S. Bukovsky, and L. O. Mearns, 2018: Pacific sea surface temperature related influences on North American monsoon precipitation within North American Regional Climate Change Assessment Program models. Int. J. Climatol., 38, 4189–4210, https://doi.org/10.1002/joc.5561.
Chen, R. D., Z. P. Wen, and R. V. Lu, 2018: Interdecadal change on the relationship between the mid-summer temperature in South China and atmospheric circulation and sea surface temperature. Climate Dyn., 51, 2113–2126, https://doi.org/10.1007/s00382-017-4002-5.
Daubechies, I., 1992: Ten Lectures on Wavelets. Society for Industrial and Applied Mathematics, 377 pp.
Ding, H., and A. Elmore, 2015: Spatio-temporal patterns in water surface temperature from Landsat time series data in the Chesapeake Bay, U.S.A. Remote Sens. Environ., 168, 335–348, https://doi.org/10.1016/j.rse.2015.07.009.
Dokulil, M. T., 2014: Impact of climate warming on European inland waters. Inland Waters, 4, 27–40, https://doi.org/10.5268/IW-4.1.705.
Glantz, M. H., 2000: Currents of Change: El Niño and La Niña Impacts on Climate and Society. 2nd ed. Cambridge University Press, 268 pp.
Kędra, M., and Ł. Wiejaczka, 2018: Climatic and dam-induced impacts on river water surface temperature: Assessment and management implications. Sci. Total Environ., 626, 1474–1483, https://doi.org/10.1016/j.scitotenv.2017.10.044.
La, B., D. Bian, Z. Ci, Z. Laba, and T. Chen, 2012: Study on the change of lake area and its causes in the Mapangyong Co Basin in Tibet (in Chinese). Arid Zone Res., 29 (6), 992–996.
Li, H. M., 2016: Research on Applications of Wavelet Transform in Signal De-nosing. 6th International Conference on Machinery, Materials, Environment, Biotechnology and Computer (MMEBC 2016), Atlantic Press, 101–105.
Liu, X. L., 2012: Trend Identification of Hydrological Sequence Based on Non-stationary and Wavelet Analysis (in Chinese). Huazhong University of Science & Technology, 68 pp.
Moser, K. A., and Coauthors, 2019: Mountain lakes: Eyes on global environmental change. Global Planet. Change, 178, 77–95, https://doi.org/10.1016/j.gloplacha.2019.04.001.
Nanjing Institute of Geography and Limnology, 2015: Atlas of Lakes Distribution in China (in Chinese). China Scientific Books, 184 pp.
NASA, 2015: Study: Climate change rapidly warming world’s lakes. NASA, https://climate.nasa.gov/news/2378/study-climate-change-rapidly-warming-worlds-lakes/.
Praveen, K., and F. G. Efi, 1997: Wavelet analysis for geophysical applications. Rev. Geophys., 35, 385–412, https://doi.org/10.1029/97RG00427.
Ptak, M., M. Sojka, T. Kałuża, A. Choiński, and B. Nowak, 2019: Long-term water temperature trends of the Warta River in the years 1960–2009. Ecohydrol. Hydrobiol., 19, 441–451, https://doi.org/10.1016/j.ecohyd.2019.03.007.
Qin, B., G. Zhu, G. Gao, Y. Zhang, W. Li, H. W. Paerl, and W. W. Carmichael, 2010: A drinking water crisis in Lake Taihu, China: Linkage to climatic variability and lake management. Environ. Manage., 45, 105–112, https://doi.org/10.1007/s00267-009-9393-6.
Sang, Y.-F., 2013: A review on the applications of wavelet transform in hydrology time series analysis. Atmos. Res., 122, 8–15, https://doi.org/10.1016/j.atmosres.2012.11.003.
Sang, Y.-F., Z. Wang, and C. Liu, 2013: Discrete wavelet-based trend identification in hydrologic time series. Hydrol. Processes, 27, 2021–2031, https://doi.org/10.1002/hyp.9356.
Schneider, P., S. J. Hook, R. G. Radocinski, G. K. Corlett, G. C. Hulley, S. G. Schladow, and T. E. Steissberg, 2009: Satellite observations indicate a rapid warming trend for lakes in California and Nevada. Geophys. Res. Lett., 36, L22402, https://doi.org/10.1029/2009GL040846.
Shirvani, A., S. M. J. Nazemosadat, and E. Kahya, 2015: Analyses of the Persian Gulf sea surface temperature: Prediction and detection of climate change signals. Arab. J. Geosci., 8, 2121–2130, https://doi.org/10.1007/s12517-014-1278-1.
Trumpickas, J., B. J. Shuter, and C. K Minns, 2009: Forecasting impacts of climate change on Great Lakes surface water temperature. J. Great Lakes Res., 35, 454–463, https://doi.org/10.1016/j.jglr.2009.04.005.
Wan, W., L. Zhao, H. Xie, B. Liu, H. Li, Y. Cui, Y. Ma, and Y. Hong, 2018: Lake surface water surface temperature change over the Tibetan Plateau from 2001 to 2015: A sensitive indicator of the warming climate. Geophys. Res. Lett., 45, 11 177–11 186 https://doi.org/10.1029/2018GL078601.
Wan, Z., S. Hook, and G. Hulley, 2015: MOD11A1 MODIS/Terra Land Surface Temperature/Emissivity Daily L3 Global 1km SIN Grid V006. NASA EOSDIS Land Processes DAAC, accessed accessed 14 January 2021, https://doi.org/10.5067/MODIS/MOD11A1.006.
Wrzesiński, D., A. Choiński, and M. Ptak, 2015: Effect of the North Atlantic Oscillation on the thermal characteristics of lakes in Poland. Acta Geophys, 63, 863–883, https://doi.org/10.1515/acgeo-2015-0001.
Wu, F., and Coauthors, 2019: Study on the change of air temperature and rainfall in the last 60 years in the Taihu watershed (in Chinese). J. Jiangsu Forest. Sci. Technol., 46, 38–42.
Wu, D. R., X. X. Geng, and M. Li, 2016: Characteristics of Climate Changes in Hulun Lake Basin in the Recent 55 Years (in Chinese). Res. Soil Water Conserv., 23, 174–184.
Xiao, F., F. Ling, Y. Du, Q. Feng, Y. Yan, and H. Chen, 2013: Evaluation of spatial-temporal dynamics in surface water temperature of Qinghai Lake from 2001 to 2010 by using MODIS data. J. Arid Land, 5, 452–464, https://doi.org/10.1007/s40333-013-0188-5.
Xu, D., Y. W. Wang, C. Jiang, P. Cheng, Y. Yan, and H. L. Huang, 2018: nalysis of the abnormal warming trend in Northeast China (in Chinese). 35th Annual Meeting of China Meteorological Society, Anhui, China, Chinese Meteor. Soc. 437–443, http://www.cms1924.org/nh35/WebPage/index.aspx.
Yan, L., and M. Zheng, 2015: The response of Lake variations to climate change in the past forty years: A case study of the northeastern Tibetan plateau and adjacent areas, China. Quat. Int., 371, 31–48, https://doi.org/10.1016/j.quaint.2014.12.057.
Yan, P. C., W. Hou, and G. L. Feng, 2016: Transition process of abrupt climate change based on global sea surface temperature over the past century. Nonlinear Processes Geophys., 23, 115–126, https://doi.org/10.5194/npg-23-115-2016.
Yang, K., and J. He, 2016: China meteorological forcing dataset (1979-2015). National Tibetan Plateau Data Center, accessed 14 January 2021, https://doi.org/10.3972/westdc.002.2014.db.
Yu, H. Y., and Coauthors, 2011: Characteristics of air temperature and precipitation in different regions of China from 1951 to 2009 (in Chinese). J. Meteor. Environ., 27, 1–11, https://doi.org/10.3969/j.issn.1673-503X.2011.04.001.
Yu, Y., Z. H. Ren, and X. Y. Meng, 2018: Construction and climate variation analysis of icing weather phenomenon series over China (in Chinese). Plateau Meteor., 37, 553–559, https://doi.org/10.7522/j.issn.1000-0534.2017.00071.
Zahidul, I., 2018: Classification of El Niño and La Niña years for water resources management in Alberta. Can. J. Civil Eng., 45, https://doi.org/10.1139/cjce-2017-0707.
Zhang, H., 2018: Long-term sequence analysis of climate factors based on wavelet analysis (in Chinese). M.S. thesis, Dept. of Mathematics, China University of Geosciences, 69 pp.
Zhang, H. X., J. Li, Q. Lin, J. Z. Qu, and Q. Z. Yang, 2013: Contrast of time-frequency analysis methods and fusion of Wigner-Ville distribution and wavelet transform. Adv. Materials Res., 805–806, 1962–1965, https://doi.org/10.4028/www.scientific.net/AMR.805-806.1962.
Zhang, Y. C., and X. F. Meng, 2005: The causes of El Nino and La Nina and their impacts on China’s climate (in Chinese). J. Chengde Teach. Coll. Natl., 25, 59–60, https://doi.org/10.16729/j.cnki.jhnun.2005.02.032.
Zhang, Z. T., and Coauthors, 2014: Application of wavelet transform to extracting the time series feature (in Chinese). Eng. Surveying Mapping, 23, 21–26.
Zhao, K. C., Y. F. Zhang, and X. Q. Zhu, 2017: NDVI changes in Ebinur basin from 2000 to 2014 and its response relation to precipitation and temperature (in Chinese). Guangdong Agric. Sci., 44, 82–88.
