The Spatiotemporal Distribution of Flash Floods and Analysis of Partition Driving Forces in Yunnan Province

Junnan Xiong, Chongchong Ye, Weiming Cheng, Liang Guo, Chenghu Zhou, Xiaolei ZhangSustainability 2019 · DOI

TL;DR

This study investigates the historical trends, spatial patterns, and driving forces of flash floods in Yunnan Province, China, revealing an increasing frequency over time and identifying key factors like precipitation and elevation. The research also assesses the sensitivity of flash floods to economic development, offering crucial insights for disaster prevention and sustainable planning.

What Problem Does This Paper Address?

Flash floods are a significant natural disaster with severe economic and societal impacts. While global research exists, understanding flash flood dynamics in complex, diverse regions like Yunnan Province, characterized by varied land cover, topography, and human activity, remains a challenge. Previous studies often treated regions uniformly, failing to account for subregional differences in land cover (e.g., grassland, settlement, farmland, forest) and their unique responses to flash flood drivers. Furthermore, the direct impact of flash floods on economic development and the nuanced interplay of multiple driving factors across these diverse land cover types needed deeper exploration.

Methods

The researchers employed a multi-faceted approach to analyze flash flood spatiotemporal patterns and driving forces:

  • Temporal Analysis: Changepoint analysis was used to identify significant shifts in annual and monthly flash flood occurrences over time.
  • Spatial Analysis:
    • Kernel Density Estimation: To visualize the spatial intensity and concentration of flash floods.
  • Spatial Mismatch Analysis: To quantify the imbalance in flash flood occurrences between different time periods.
  • Standard Deviational Ellipse (SDE): To delineate the directional trend and dispersion of flash flood distribution.
  • Spatial Gravity Center Model: To track the migration of flash flood centers, precipitation, and population over time.
  • Driving Force Analysis:
    • Pearson Correlation Coefficient: To assess the linear relationship between flash floods and various driving factors (precipitation, elevation, topography, land cover indices, population density, GDP) within different land cover subregions.
  • Principal Component Analysis (PCA) and Multiple Linear Regression: To explore the interaction and relative importance of multiple driving factors, addressing multicollinearity and identifying dominant influences.
  • Sensitivity Analysis: A quantitative model was used to assess the sensitivity of flash floods to economic development (GDP) across different time periods and regions.

Datasets

The study utilized the following datasets:

  • Flash Flood Data: National Flash Flood Investigation and Evaluation Project (1949-2015).
  • Precipitation Data: National Flash Flood Investigation and Evaluation Project (20-year return period for 10 min, 60 min, 6 h, and 24 h precipitation), and annual mean precipitation data from the National Weather Service of China.
  • Digital Elevation Model (DEM): China Geospatial Data Cloud (2000).
  • Population Density Data: Resources and Environmental Sciences Data Center (2000, 2005, 2010, 2015).
  • Gross Domestic Product (GDP) Data: Resources and Environmental Sciences Data Center (1995, 2000, 2005, 2010, 2015).
  • Normalized Difference Vegetation Index (NDVI): NASA's MODIS (2001-2015).
  • Landcover Data: China Global Change Research Data Publishing & Repository (2010).

Key Insights and Ideas

  • Increasing Trend: Flash floods are not static; their annual frequency has been steadily increasing, highlighting a growing hazard.
  • Subregion Matters: The impact of driving factors varies significantly across different land cover types. What's crucial for flash floods in settlements might be less so in grasslands, emphasizing the need for localized analysis.
  • Precipitation is King, but Elevation is a Close Second (in Settlements): While precipitation is a universal driver, elevation plays a surprisingly strong role in settlements, suggesting that human infrastructure in higher-altitude areas is particularly vulnerable.
  • Economic Development's Double-Edged Sword: While economic development can increase sensitivity to flash floods, particularly in rapidly developing areas like Kunming, long-term sustainable policies can help mitigate this.
  • Short-Term vs. Long-Term Strategies: The findings suggest a clear distinction: focus on heavy precipitation for immediate forecasting, but address human activities and ecosystem vulnerability for long-term resilience.

Key Results

  • Temporal Trend: The number of flash floods in Yunnan increased gradually from 1949 to 2015. The period 1996-2015 saw the highest frequency, with an average of 102 flash floods per year. Monthly peaks occurred in June, July, and August.
  • Spatial Hotspots: Regions with high flash flood occurrences were concentrated in Zhaotong, Qujing, Kunming, Yuxi, Chuxiong, Dali, and Baoshan.
  • Driving Factors in Settlements: Precipitation (10 min, 60 min, 24 h, 6 h return periods) and elevation showed the strongest correlations with flash floods in settlement areas, with Pearson's r values ranging from 0.395 to 0.675.
  • Sensitivity to Economy: Kunming exhibited the highest sensitivity (S = 21.86) to economic development during 2000-2005, indicating a strong link between economic activity and flash flood impacts in that period. However, sensitivity generally decreased in later periods.

Limitations and Future Work

The authors acknowledge several limitations:

  • Data Availability: The time series for some driving factors (e.g., GDP, population density) did not extend back to 1949, limiting the scope of long-term analysis for these variables.
  • Precipitation Interpolation: The use of standard interpolation methods for precipitation might not fully capture the nuances of mountain terrain, potentially affecting accuracy.
  • Scale of Analysis: While the spatial analysis methods (kernel density, SDE, etc.) are effective for large-scale regional analysis, their application to small-scale features like flood protection works or local geomorphology is limited.
  • Principal Component Analysis: The method for determining the number of principal components can be subjective and might lead to the loss of local information.

Future work could involve:

  • Extending the time series of driving factor data.
  • Employing more sophisticated precipitation interpolation techniques that account for complex terrain.
  • Developing or applying methods suitable for finer-scale flash flood analysis.
  • Refining PCA methods to better preserve local information.

Summary updated

Abstract

Flash floods are one of the most serious natural disasters, and have a significant impact on economic development. In this study, we employed the spatiotemporal analysis method to measure the spatial–temporal distribution of flash floods and examined the relationship between flash floods and driving factors in different subregions of landcover. Furthermore, we analyzed the response of flash floods on the economic development by sensitivity analysis. The results indicated that the number of flash floods occurring annually increased gradually from 1949 to 2015, and regions with a high quantity of flash floods were concentrated in Zhaotong, Qujing, Kunming, Yuxi, Chuxiong, Dali, and Baoshan. Specifically, precipitation and elevation had a more significant effect on flash floods in the settlement than in other subregions, with a high r (Pearson’s correlation coefficient) value of 0.675, 0.674, 0.593, 0.519, and 0.395 for the 10 min precipitation in 20-year return period, elevation, 60 min precipitation in 20-year return period, 24 h precipitation in 20-year return period, and 6 h precipitation in 20-year return period, respectively. The sensitivity analysis showed that the Kunming had the highest sensitivity (S = 21.86) during 2000–2005. Based on the research results, we should focus on heavy precipitation events for flash flood prevention and forecasting in the short term; but human activities and ecosystem vulnerability should be controlled over the long term.
BibTeX
@inproceedings{xiong2019spatiotemporal,
  author = {Junnan Xiong and Chongchong Ye and Weiming Cheng and Liang Guo and Chenghu Zhou and Xiaolei Zhang},
  title = {{The Spatiotemporal Distribution of Flash Floods and Analysis of Partition Driving Forces in Yunnan Province}},
  booktitle = {Sustainability},
  year = {2019},
  doi = {10.3390/su11102926},
}
Figure