A lake data set for the Tibetan Plateau from the 1960s, 2005, and 2014

Wei Wan, Di Long, Yang Hong, Yingzhao Ma, Yuan Yuan, Pengfeng Xiao, Hongtao Duan, Zhongying Han, Xingfa GuScientific 2016 · DOI

TL;DR

This paper introduces a comprehensive, freely available dataset of Tibetan Plateau lakes from the 1960s, 2005, and 2014. This dataset, derived from historical surveys and high-resolution satellite imagery, addresses a critical need for understanding the region's environmental changes and the complex interactions within its cryosphere, hydrosphere, and atmosphere.

What Problem Does This Paper Address?

The Tibetan Plateau (TP), often referred to as the "third pole," is a crucial region for understanding global climate and hydrological processes. Lakes on the TP are vital components of its ecosystem, influencing biogeochemistry and acting as indicators of environmental change. However, the harsh environment and limited accessibility of the TP have historically resulted in a scarcity of comprehensive, long-term data on its lakes. Existing global datasets often lack the necessary resolution or temporal coverage for detailed regional studies. This paper aims to fill this gap by providing a consistent, multi-temporal dataset of lake characteristics, enabling more robust research into the TP's environmental dynamics.

Methods

The researchers employed a multi-pronged approach to construct their lake dataset across three distinct time periods:

  • 1960s: This dataset was derived from historical ground surveys and mapping efforts conducted as part of China's first nationwide lake investigation. The data was originally in a 1:250,000 scale ESRI shapefile format. Minor edits were made to ensure lake uniqueness and attribute consistency.
  • 2005: Lake data for this period was primarily sourced from China-Brazil Earth Resources Satellite (CBERS-1) Charge-Coupled Device (CCD) sensor images (19.5 m spatial resolution). Landsat Enhanced Thematic Mapper Plus (ETM+) images were used as supplementary data to fill gaps due to cloud cover. Images were geometrically corrected and geo-rectified.
  • 2014: High-resolution images from China's newly launched GaoFen-1 (GF-1) Wide Field of View (WFV) cameras (16 m spatial resolution) formed the basis for this dataset. Landsat 8 Operational Land Imager (OLI) images were also utilized. GF-1 images were ortho-rectified prior to analysis.

A key methodological choice was the manual interpretation and digitization of satellite imagery for water boundary extraction. This approach was chosen over automatic methods to ensure higher accuracy and reliability, especially given the potential uncertainties in automated algorithms. Specific rules were developed to handle different lake types (freshwater, semi-saline, saline), formation mechanisms (tectonic, barrier, fluvial), and specific issues like islands, bottomlands, and merging lakes. New-born and dead lakes were also identified based on their presence or absence across different time periods.

Datasets

The primary datasets used in this study are:

  • 1960s Lake Survey Data: Historical ground survey and mapping data (1:250,000 scale).
  • 2005 Satellite Imagery:
    • China-Brazil Earth Resources Satellite (CBERS-1) CCD images.
  • Landsat Enhanced Thematic Mapper Plus (ETM+) images.
  • 2014 Satellite Imagery:
    • China's GaoFen-1 (GF-1) Wide Field of View (WFV) cameras.
  • Landsat 8 Operational Land Imager (OLI) images.

The study also references and compares its findings with existing global datasets such as the Global Lakes and Wetlands Database (GLWD) and a regional dataset by Yao et al. (2000) for the Hoh Xil region.

Key Insights and Ideas

  • The critical role of manual interpretation: The paper strongly advocates for manual interpretation in lake boundary extraction for high-accuracy regional studies, especially in complex environments like the TP, where automatic methods might falter.
  • Bridging temporal data gaps: The creation of a three-time-series dataset (1960s, 2005, 2014) is a significant contribution, allowing for the study of lake dynamics over a substantial period.
  • Understanding lake evolution: The identification of "new-born" and "dead" lakes provides direct evidence of dynamic environmental changes, such as lake expansion and shrinkage, driven by climatic and hydrological shifts.
  • Regional variations in lake change: The data reveals distinct trends across different basins, with some areas showing consistent expansion (e.g., Inner TP basin) and others showing contraction (e.g., Brahmaputra basin), highlighting the spatially heterogeneous impacts of climate change.

Key Results

  • The study presents a comprehensive dataset covering lakes with areas greater than 1 km² across the Tibetan Plateau for three distinct time points.
  • The dataset includes attributes such as lake name, location, shape, area, and perimeter.
  • Statistical analysis shows a general trend of increasing lake area across the TP from the 1960s to 2014.
  • The Inner TP basin exhibited a significant average growth rate of 9.88% between 2005 and 2014, with specific sub-basins showing even higher rates.
  • Conversely, lakes in the Brahmaputra basin showed a decreasing rate of -2.53% during the same period.
  • Validation against other datasets (GLWD, Yao et al.) indicates good consistency, reinforcing the reliability of the developed dataset.

Limitations and Future Work

  • Exclusion of small lakes: The dataset focuses on lakes with areas ≥ 1 km², excluding smaller lakes (< 1 km²) which can also be significant for biogeochemical processes. The authors encourage future work to include these smaller water bodies.
  • Limited temporal resolution: While three time points are provided, more frequent data would be beneficial for a more detailed understanding of rapid environmental changes.
  • Manual interpretation workload: The authors acknowledge that automatic methods are more efficient and suggest future work to compare manual interpretation with advanced automatic methods.
  • Need for more data: The authors emphasize that to study detailed information on change characteristics and to investigate particular lake or basin-scale water balance, data acquired at more time intervals and effective automatic methods are required. They welcome contributions to expand the dataset with smaller lakes, new time series, and additional attributes.

Summary updated

Abstract

Long-term datasets of number and size of lakes over the Tibetan Plateau (TP) are among the most critical components for better understanding the interactions among the cryosphere, hydrosphere, and atmosphere at regional and global scales. Due to the harsh environment and the scarcity of data over the TP, data accumulation and sharing become more valuable for scientists worldwide to make new discoveries in this region. This paper, for the first time, presents a comprehensive and freely available data set of lakes' status (name, location, shape, area, perimeter, etc.) over the TP region dating back to the 1960s, including three time series, i.e., the 1960s, 2005, and 2014, derived from ground survey (the 1960s) or high-spatial-resolution satellite images from the China-Brazil Earth Resources Satellite (CBERS) (2005) and China's newly launched GaoFen-1 (GF-1, which means high-resolution images in Chinese) satellite (2014). The data set could provide scientists with useful information for revealing environmental changes and mechanisms over the TP region.
BibTeX
@inproceedings{wan2016lake,
  author = {Wei Wan and Di Long and Yang Hong and Yingzhao Ma and Yuan Yuan and Pengfeng Xiao and Hongtao Duan and Zhongying Han and Xingfa Gu},
  title = {{A lake data set for the Tibetan Plateau from the 1960s, 2005, and 2014}},
  booktitle = {Scientific},
  year = {2016},
  doi = {10.1038/sdata.2016.39},
}
Figure