Zone II • Versions EN1
Abstract: As an ecologically sensitive region with the highest altitude, the Tibetan Plateau has been a hotspot in the study of long-term climate changes. Extreme weather events are related to the regional climate change and are an important component of the study. Based on the daily temperature and precipitation observed at 99 meteorological stations over the Tibetan Plateau during 1960 – 2012, this study calculated a series of climate extreme indices. Firstly, quality control was performed on the raw meteorological observations to eliminate abnormal values. Then the R package RClimDex1.0 was used to compute 15 temperature and 8 precipitation extreme indices. 10 indices were determined by the absolute observed value and 30 indices determined by the threshold approach. This dataset is useful for research on the occurrence frequency and developing trend of extreme events. When used together with regional statistical data, this dataset can also reveal the impacts of climate extremes on agricultural and livestock production.
Keywords: Tibetan Plateau; climate extreme indices; climate change; meteorological stations
|Chinese title||1960 ～ 2012 年青藏高原极端气候指数数据集|
|English title||A dataset of climate extreme indices over the Tibetan Plateau (1960 – 2012)|
|Corresponding author||Zhou Yuke (email@example.com)|
|Data authors||Zhou Yuke, Gao Qi|
|Time range||1960 – 2012|
|Geographical scope||The Tibetan Plateau (26°00'N – 39°47'N, 73°19'E – 104°47'E), including 99 meteorological stations|
|Temporal resolution||Year||Data volume||2.39 MB|
|Data service system||<http://www.sciencedb.cn/dataSet/handle/444>|
|Sources of funding||National Natural Science Foundation of China (2016, Grant No. 41601478);|
Sciences Technology Service Project of the Chinese Academy of Sciences (2016, Grant No. KFJ-SW-STS-167)
|Dataset composition||This dataset consists of 23 indices documents in CSV format, including 15 extreme temperature indices and 8 extreme precipitation indices. Files of this dataset are named after their index names in English. Each file consists of year, inter- annual index results, smoothing results and site identifier. Each file has a slightly different data volume depending on the number of sites and the year of observation. The amount of uncompressed data is about 80 – 160 KB.|
Under the impact of increasing human socio-economic activities and global climate changes,1-2 catastrophic weather events occurred frequently, such as global heat wave weather, extreme precipitation, tropical storms and sea level changes.3 Due to the Tibetan Plateau's sensitivity to northern hemisphere and even global climate changes as well as its own ecological vulnerability,4–5 scholars have carried out extensive studies on climate development, trends and vegetation responses in this alpine and arid region.6–7
Analysis of climate change requires ground meteorological data as its basis. Despite the comprehensive surface observation data provided by the China Meteorological Science Data Sharing Service Network which include 16 parameters covering precipitation, wind speed, water vapor pressure, relative humidity, sunshine and air temperature, it cannot be used directly to analyze climate changes because of its large data amount and data redundancy caused by some of its uncommon parameters. Based on the above meteorological data, it is possible to calculate a series of extreme climate indices. Proposed by the World Meteorological Organization (WMO), these indices provide a unified standard and basis for the statistics and analysis of temperature and precipitation events. They not only straightforwardly reflect the distribution and occurring trend of extreme climate events,8 but also provide a reference for the region's agriculture and animal husbandry development and ecological environment management.9 Data of extreme weather indices are therefore important as it supports climate research in the Tibetan Plateau.
In this article, the dataset of extreme climate indices includes absolute and relative indices about extreme temperature and precipitation over the Tibetan Plateau, which can reflect the region's inter-annual temperature and precipitation. The indices describe in detail the cold or warm temperature, the extremum of low or high temperature, the length of vegetation growth season, the number of consecutive cold or warm days, the annual precipitation, the heavy precipitation, and the number of consecutive wet or dry days. They provide data support for further analyzing the inter-annual variation and changing characteristics of extreme temperature and precipitation in long time series over the Tibetan Plateau.
2.1 Meteorological stations distribution
The raw data used in this study are from the Meteorological Data Sharing Service Network of China. The meteorological stations are mainly distributed in middle and eastern Tibetan Plateau, with an elevation of 2,500 to 4,500m that enables a coverage of Xinjiang Uygur Autonomous Region, Qinghai Province, Gansu Province, Sichuan Province, Yunnan Province and Tibet Autonomous Region. They span three temperature zones – plateau temperate zone, sub-frigid zone and subtropical zone, and more than 10 types of ecogeographic regions. Among the 106 meteorological stations, seven did not generate observation data, and the remaining 99 sites were effective for this study. Figure 1 shows the distribution of the meteorological stations on the Tibetan Plateau and their spatial elevation.
2.2 Data production process
The data production process is divided into four parts: data preprocessing, data loading and quality control, calculation of extreme climate indices, and indices results sorting. The overall process is shown in Figure 2.
2.2.1 Data preprocessing
The raw data are from the Meteorological Data Sharing Service System of China. Firstly, independent data files for the same year from all the sites are merged; Secondly, six data items are extracted for RClimDex1.0 to calculate extreme climate indices, including year, month, day, daily maximum temperature (Tmax), daily minimum temperature (Tmin) and daily precipitation (Prcp) of all the stations. Finally, save them as a CSV file. The above-mentioned software is an R language script which comes from the Climate Research Branch, Environment Canada.10
The observed raw values of precipitation are made up of codes of precipitation type and volume. The first two digits represent precipitation type, and 31, 32, 30 refer to snowfall, fog and dew, sleet respectively. The latter three digits stand for effective precipitation values. Among them, the special value 32700 represents a precipitation volume of less than 0.1 mm, which can be replaced by 0 because it has little effect on life and production. Moreover, 32766 is a missing value, 32744 is a null value, and both of them can be replaced by –99.9, a format recognizable by the software.
The unit of the recorded raw temperature value is 0.1°C, and it needs to be decoded into an actual value according to the relationship between the recorded value and the actual temperature (Equation 1). In addition, the missing values in the temperature records are likewise replaced by –99.9.
After Tmax, Tmin and Prcp are decoded, all the records of each site are extracted and ordered by date. Lastly, output text files in ASCII format named after 99 stations, which helps meet the needs of subsequent calculations of extreme climate indices.
2.2.2 Extreme climate indices calculation
Before calculating the extreme climate indices, it is necessary to set the parameters such as the beginning and end year of the base period, the latitude and longitude of the meteorological stations, the upper and lower limits of daily maximum or minimum temperature and the precipitation threshold. Among them, the beginning and end year can be used to calculate the base period, while one record with missing data should be added in front of the original data to ensure the integrity of time series. The upper limit of daily maximum temperature can be used to calculate the warm consecutive index, while the lower limit of the minimum temperature can be used to calculate the cold consecutive index. Additionally, it is considered to be consecutive wet days when the daily precipitation in the raw data is greater than the set threshold. This study selects and calculates 23 typical extreme climate indices, including 15 temperature indices and 8 precipitation indices.
2.2.3 Data smoothing
In order to eliminate the instability of time series of all the indices, a simple sliding averaging method (SMA) is used to perform data smoothing, which takes the average value of consecutive five years as the smoothing result. Because there are too many enumerations for the extreme indices, we take the CDD (Consecutive Dry Days) index as an example. The data smoothing process is shown in Figure 3.
The workflow of data smoothing is described as follows: firstly, read the file of index results and store them as f1, which consists of years, corresponding interannual index results and site ID; secondly, extract all records identified by the first station ID and store them as f, then convert the data of interannual index to time series (×1) and make ×1 simple smoothing; f and the smoothing result SMA are then merged to fa as the basis of subsequent site result merging; thirdly, iterative loop is used to process matrix fi identified by the i-th (1<i≤ns, ns represents the number of sites) sites name ID, and repeat the smoothing process of the first site data. After merging the i-th smoothing results and fi into fb, the merging result fa with the last iteration is merged into a new fa. Until i is equal to ns, the iterative smoothing process is complete; fourthly, when i appears to be greater than ns, output the merging result fa in CSV format, for the convenience of analyzing and applying extreme temperature and precipitation over the Tibetan Plateau.
The dataset contains 23 typical extreme climate indices results of 99 sites on the Tibetan Plateau. For the convenience of calculation and application, the dataset is stored in CSV format, which includes 15 temperature indices and 8 precipitation indices. The results are named after English names of these extreme indices, such as CDD.CSV. Table 1 shows all the indices of this dataset and their explanation.
|TN10p||Number of days when daily minimum temperature < 10th percentile||d|
|2||TN90p||Number of days when daily minimum temperature > 90th percentile||d|
|3||TX10p||Number of days when daily maximum temperature < 10th percentile||d|
|4||TX90p||Number of days when daily maximum temperature > 90th percentile||d|
|5||TXx||Maximum value of yearly maximum temperature||℃|
|6||TNx||Maximum value of yearly minimum temperature||℃|
|7||TXn||Minimum value of yearly maximum temperature||℃|
|8||TNn||Minimum value of yearly minimum temperature||℃|
|9||TMAXmean||Mean value of yearly maximum temperature||℃|
|Mean value of yearly minimum temperature||℃|
|11||FD0||Number of days when daily minimum temperature < 0ºC within one year||d|
|12||WSDI||Number of days when daily maximum temperature > 90th percentile during six consecutive days||d|
|13||CSDI||Number of days when a continuous 6d when TN<10th percentile||d|
|14||GSL||Continuous 6d> 5 ºC or <5 ºC time span||d|
|15||SU25||Days of daily maximum temperature > 25 ºC||d|
|16||Precipitation indices||SDII||Precipitation divided by wet days in the year||mm/d|
|17||R10||Annual count of days when precipitation >=10mm||d|
|18||CWD||the longest consecutive days when daily precipitation <1mm||d|
|19||CDD||the longest consecutive days when daily precipitation >1mm||d|
|20||R95p||Annual total precipitation when precipitation >95th percentile||mm|
|21||RX5day||Monthly maximum precipitation for a continuous 5d||mm|
|22||RX1day||Monthly maximum precipitation for 1d||mm|
|23||PRCPTOT||Annual total precipitation in wet days (precipitation >=1mm)||mm|
Among them, each index result file contains four columns of attribute values: year, interannual index value, smoothing value, and corresponding site ID (e.g., 51804).
Quality control was conducted from two aspects: ① during the preprocessing process, special values in the raw data were interpreted and data was unified in the common unit; ② in the process of extreme climate indices production, quality control was strictly performed on the preprocessed data through combination of manual inspection and software automatic monitoring. Before extreme climate indices were calculated, data outliers in each site file were detected using the RClimDex1.0 software, which generated the result files of quality control. Data were then corrected or deleted after being validated through manual inspection.
4.1 Preliminary quality control
Data quality control is a prerequisite of the extreme climate indices calculation. The outliers and error values of the raw data not only causes wrong results, but also affects subsequent trend analysis. The RClimDex1.0 software package in R language environment is used to process meteorological observations, including quality control of extreme temperature and precipitation data and calculation of extreme climate indices.
Firstly of all, data files of all the meteorological stations in TXT format are checked whether there are any illogical outliers. The contents of the inspection include: ① the maximum daily temperature (Tmax) is less than the daily minimum temperature (Tmin); ② the precipitation (Prcp) is less than 0 mm. Use RClimDex1.0 software to load the preprocessed data, run the module of quality control function, and the software automatically identifies –99.9 for NA and replaces abnormal values with NA. Secondly, in order to detect outliers in the temperature and precipitation time series, reasonable values are defined as not more or less than three times the daily mean value of the climate time series over the Tibetan Plateau, and the outliers are expressed as [–∞, Mean – 3*Std] U [Mean+3*Std, +∞]. If the original data are within the outlier range [–∞, Mean–3*Std] U [Mean+3*Std, +∞], the records in the quality control result files must be checked and screened again. Eliminate unreasonable records or set them as missing values to ensure that data quality is strictly controlled; extreme climate indices can be calculated only when there are no abnormal outliers.
Based on the statistical analysis of the observations, it is further checked whether there are unreasonable outliers. Figure 4 shows the probability density distribution of the temperature and precipitation observations at site 51804. As 95% of the distribution out of the distribution area are likely to be abnormal values, we checked the raw data in this circumstance to decide whether to remove.
4.2 Evaluation of extreme climate indices products
In order to intuitively display the variation trend reflected by the extreme climate indices, the variation slope of typical extreme indices of each station is expressed by contours, which can further verify the reliability of the results of this dataset. Figures 5 and 6 show the visualization results of typical temperature and precipitation indices in each station respectively, where a represents the interannual trend of each index every 10 years and the unit is d/10a.
According to the trend as reflected by the long time series of temperature and precipitation indices in the above figure, it can be seen that the variation slope of the cold index is negative for most of the sites while that of the warm index is mostly positive, and that of the precipitation index is smaller positive. This conforms to the background of global warming, and is consistent with the changes of temperature and precipitation in the previous study over the Tibetan plateau.
The frequency and variation trend of extreme climate events have a direct impact on the analysis and evaluation of climate change. The dataset can be used in conjunction with conventional meteorological observation data to explore the trend and spatial characteristics of long-term climate changes over the Tibetan Plateau. For instance, extreme temperature indices consist of cold index and warm index, and an analysis of the two indices' changing trend is conducive to understanding temperature differentiation. It can also be used to correlate with local socio-economic statistics to assess the impact of extreme climate events on agricultural production.
The dataset shares the results of 23 extreme climate indices observed by 99 meteorological stations over the Tibetan Plateau from 1960 to 2012. Files are stored by indices in CSV format for the convenience of subsequent processing and application. Users can download the data as needed.
Zhou Yuke, PhD, Assistant Professor; research area: ecological remote sensing. Contribution: design of data product and implementation of key technology.
Gao Qi, Master’s student; research area: ecological remote sensing. Contribution: calculation of extreme climate indices.
We thank Dr. Zhang Wenjie from the State Key Laboratory of Resources and Environmental Information Systems for his guidance on data processing and Dr. Fan Junfu who provides the experimental environment.
Li H & Li L. Mean and climate change on the Qinghai-Tibetan Plateau with a 2℃ global warming. Progressus Inquisitions De Mutatione Climatis 11 (2015): 157 – 164.
Alexander L, Zhang X, Peterson T et al. Global observed changes in daily climate extremes of temperature and precipitation. Journal of Geophysical Research 111 (2006): 1042 – 1063.
Zwiers F, Alexander L, Hegerl G et al. Climate extremes: Challenges in estimating and understanding recent changes in the frequency and intensity of extreme climate and weather events, in Climate Science for Serving Society, 339 – 389, ed. Ghassem R & James W. Netherlands: Springer, 2013.
Wang B, Bao Q, Hoskins B et al. Tibetan Plateau warming and precipitation changes in East Asia. Geophysical Research Letters 35 (2008), doi: 10.1029/2008GL034330.
Yao Y & Zhang B. The spatial pattern of monthly air temperature of the Tibetan Plateau and its implications for the geo-ecology pattern of the Plateau. Geographical Research 34 (2015): 2084 – 2094.
Wu S, Yin Y, Zheng D et al. Climate changes in the Tibetan Plateau during the last three decades. Acta Geographica Sinica 60 (2005): 3 – 11.
Wu G, Duan A, Zhang X et al. Extreme weather and climate changes and its environmental effects over the Tibetan Plateau. Chinese Journal of Nature 35 (2013): 167 – 171.
You Q, Kang S, Aguilar E et al. Changes in daily climate extremes in the eastern and central Tibetan Plateau during 1961 – 2005. Journal of Geophysical Research: Atmospheres 113 (2008), doi:10.1029/2007JD009389.
Zhou B, Xu Y, Wu J et al. Changes in temperature and precipitation extreme indices over China: Analysis of a high-resolution grid dataset. International Journal of Climatology 36 (2016): 1051 – 1066.
1. Zhou Y & Gao Q. A dataset of climate extreme indices over the Tibetan Plateau (1960 – 2012). Science Data Bank. DOI: 10.11922/sciencedb.444
How to cite this article
Zhou Y & Gao Q. A dataset of climate extreme indices over the Tibetan Plateau (1960 – 2012). China Scientific Data 2 (2017). DOI: 10.11922/ csdata.170.2017.0143