Mapping Heat-Related Risks in Northern Jiangxi Province of China Based on Two Spatial Assessment Frameworks Approaches

Abstract

Heat-health risk is a growing concern in many regions of China due to the more frequent occurrence of extremely hot weather. Spatial indexes based on various heat assessment frameworks can be used for the assessment of heat risks. In this study, we adopted two approaches-Crichton's risk triangle and heat vulnerability index (HVI) to identify heat-health risks in the Northern Jiangxi Province of China, by using remote sensing and socio-economic data. The Geographical Information System (GIS) overlay and principal component analysis (PCA) were separately used in two frameworks to integrate parameters. The results show that the most densely populated community in the suburbs, instead of city centers, are exposed to the highest heat risk. A comparison of two heat assessment mapping indicates that the distribution of HVI highlights the vulnerability differences between census tracts. In contrast, the heat risk index of Crichton's risk triangle has a prominent representation for regions with high risks. The stepwise multiple linear regression zero-order correlation coefficient between HVI and outdoor workers is 0.715, highlighting the vulnerability of this particular group. Spearman's rho nonparametric correlation and the mean test reveals that heat risk index is strongly correlated with HVI in most of the main urban regions in the study area, with a significantly lower value than the latter. The analysis of variance shows that the distribution of HVI exhibits greater variety across urban regions than that of heat risk index. Our research provides new insight into heat risk assessment for further study of heat health risk in developing countries.

Keywords: Crichton’s risk triangle; developing countries; heat vulnerability index (HVI); heat-health risk; spatial risk assessment.

Figure A1

Spearman’s rho nonparametric correlations and significances among the selected variables used in spatial analysis (significance levels: **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).).

Figure A2

Scree plots and loading factor plots of (a) and (b) population variables; (c) and (d) heat vulnerability and adaptation variables.

Figure 1

(a) elevation and (b) land cover type of the study area. Elevation data source: ASTER GDEM (2009), ERSDAC.

Figure 2

The number of days with temperature above 35 °C in Nanchang from 1979 to 2018. Air temperature data source: National meteorological information center.

Figure 3

A detailed flowchart of spatial risk assessment methodology.

Figure 4

(a) daytime land surface temperature (LST); (b) nighttime LST; (c) the annually average number of days with the highest temperature >35 °C in each county from 2015 to 2018; (d) the annually average number of days with the highest air quality index (AQI) level in each county in the summer from 2015 to 2018; (e) spatial distribution of 8-day average wet bulb globe temperature (WBGT); (f) spatial distribution of heat hazard/exposure index.

Figure 5

Comparison of total population of each county in Landscan and statistical yearbook. The horizontal axis is the total population at the end of 2015 (ten thousand people), and the vertical axis is the zonal statistics from Landscan after normalization to 0–1.

Figure 6

(a,c): Heat exposure index and Heat vulnerability index of Crichton’s risk triangle; (b,d): Sensitivity and Adaptability of heat vulnerability index (HVI).

Figure 7

(a) spatial distribution of heat risk under Crichton’s risk triangle framework; (b) spatial distribution of high-temperature vulnerability index under HVI framework. (c) three selected main urban regions (Nanchang, Xinyu and Yichun) and their spatial distributions of heat risk index and HVI.

Figure 8

Spearman’s rho correlation coefficients of the nine main urban regions in the study area. ** means correlation is significant at the 0.01 level (2-tailed), * means correlation is significant at the 0.05 level (2-tailed).

Figure 9

Comparing nine main urban areas (horizontal indexes 1–9 refer to Pingxiang, Yichun, Xinyu, Fuzhou, Yingtan, Shangrao, Nanchang, Jingdezhen and Jiujiang). The boxes represent the 25th to the 75th percentile; the crosses are the maximum and minimum values. The number in bracket is the quantity of urban regions that are significantly different from the urban of the corresponding horizontal index. The significance level is 0.05.