【摘要】：目的 對我國部分城市的居民死因資料和氣象資料進行關聯性分析,在此基礎上進行高溫熱浪對城市居民健康影響的模型擬合分析及初步預測分析,探索適合我國大中城市的熱浪對居民健康影響的預測模型,為預測預警熱浪的健康影響提供基礎資料和科學依據。 方法 收集哈爾濱市及北京市朝陽區(qū)的居民死因數據和氣象數據,利用滯后模型和灰色關聯等方法對兩市的數據資料進行分析,并應用分段回歸模型進行閾值分析,結合高溫期間的現場調查結果和訪談信息,以及前期對重慶市和汕頭市的研究結果,對多個城市進行廣義相加模型擬合分析及運用分布滯后非線性模型進行初步預測分析。 結果 1.現場調查數據顯示:高溫熱浪可以導致熱相關疾病尤其是中暑的發(fā)病率增加,雖然居民在高溫期間會有意識地采取一些應對措施,但中老年人群仍是易感人群。 2.典型城市的氣象因素與居民死亡指標的關系分析顯示:高溫對死亡指標的影響主要集中在滯后0至4天,混雜因素可吸入顆粒物濃度的影響有一定的滯后,其濃度每上升10μg/m3所對應的相對危險度變化在滯后10天左右達較高水平,累積相對危險度在滯后15天左右達峰值,環(huán)境溫度在滯后0天和1天具有較高的相對危險度,可吸入顆粒物濃度在200μg/m3以下時與環(huán)境溫度的交互作用較為明顯;死亡指標變化的長期趨勢與氣溫和相對濕度有關,此基礎上,氣壓是影響死亡指標短期波動的主要因素。 3.多個城市的環(huán)境溫度閾值分析顯示,北京市、哈爾濱市、重慶市、福州市及汕頭市的全死因死亡危險的閾值溫度(℃)分別為25.63(±0.809)、23.24(±1.114)、29.29(±2.886)、36.06(±0.281)和31.28(±1.016),當各城市的環(huán)境溫度分別高于相應的閾值溫度時,每升高1℃死亡人數增加的百分數(95%置信區(qū)間)分別為0.99(0.46~1.52)、1.12(0.24~2.00)、0.17(-0.34~0.68)、21.70(11.99~32.26)和2.79(1.05~4.57),其中福州市的結果明顯偏高,可能是由于數據質量問題,因此該結果有待進一步論證。 4.僅考慮溫度因素時模型擬合分析確定的各城市全死因死亡指標的預測因子分別為北京市l(wèi)ag0+lag1+lag24,重慶市l(wèi)ag0,福州市l(wèi)ag0+lag1521,汕頭市l(wèi)ag0+lag24+lag814,哈爾濱市l(wèi)ag0+lag1+lag814+lag1521,納入相對濕度因素后,各城市的預測因子分別為北京市l(wèi)ag0+lag1+lag24+lag1521,重慶市無,福州市無,汕頭市l(wèi)ag0+lag814,哈爾濱市l(wèi)ag0+lag1+lag814+lag1521。 5.初步預測分析顯示:高溫對居民死亡指標的影響主要集中在滯后0至4天左右,分布滯后非線性模型可通過溫度的變化及滯后時間的確定估計死亡指標相對危險度的變化,進而確定欲觀察時間段內的死亡指標危險程度。北京市和重慶市高溫期間對應的最大相對危險度值分別為1.126和1.118,與此相對應的溫度并不是最高日平均氣溫,這種現象主要可能是因為人群對高溫的耐受及應對極端高溫采取相應的措施,以至極端高溫對應的相對危險度有所下降;汕頭市、福州市及哈爾濱市高溫期間最大的相對危險度分別為1.310、1.269和1.254,分別對應于各城市的最高日平均氣溫。 結論 1.高溫的滯后效應主要在滯后0～4天,為急性效應,可吸入顆粒物污染在氣溫的影響過程中存在較長的滯后效應;居民死亡指標的長期趨勢與氣溫及相對濕度有關,而氣壓在氣溫和相對濕度的影響基礎上,主要影響死亡指標的短期波動。 2.僅以氣溫因素與混雜因子進行廣義相加模型擬合時,不同的城市滯后0天溫度的效應均有統(tǒng)計意義,而氣溫效應的滯后時間長短有一定的差異;將相對濕度因素納入模型時,重慶市和福州市的顯著性因子有較大變化,表明相對濕度因素在重慶市和福州市對氣溫的健康影響有明顯的混雜效應。 3.分布滯后非線性模型能夠對不同的溫度和不同滯后時間的死亡指標的相對危險度進行估計,通過觀察日的溫度相對于參考日的溫度變化及確定一個滯后時間段,了解觀察日相對于參考日的相對危險度變化,以此對觀察日的居民死亡指標的相對危險度進行預測;通過對模型的初步驗證,該模型可以用于預測高溫對健康的影響,但數據質量對模型的估計效果有較大影響。
[Abstract]:objective
On the basis of the correlation analysis of the death causes and meteorological data of some urban residents in China, the model fitting analysis and preliminary prediction analysis on the health impact of high temperature heat waves on urban residents' health are carried out to explore the prediction model of the health impact of the heat waves suitable for the large and middle cities in China, in order to predict the health effects of the early warning heat waves. For basic information and scientific basis.
Method
The data of death causes and meteorological data of residents in Harbin and Chaoyang District of Beijing city are collected. The data of the two cities are analyzed by means of lag model and grey correlation, and the threshold analysis is carried out by the piecewise regression model, and the field investigation results and interview information during the high temperature are combined, and the research on the city of Chongqing and Shantou in the early stage is also carried out. The results show that the generalized additive model fitting analysis and preliminary prediction analysis are carried out for several cities by using the distributed lag nonlinear model.
Result
1. field survey data show that high temperature heat waves can lead to an increase in the incidence of heat related diseases, especially heatstroke, although the residents will take some measures consciously during the high temperature, but the elderly are still susceptible.
The analysis of the relationship between the meteorological factors and the death index of 2. typical cities showed that the influence of high temperature on the death index was mainly concentrated in 0 to 4 days, and the influence of the concentration of inhalable particulate matter was lagging behind, and the relative risk of the concentration of its concentration was higher at about 10 g/m3 and accumulated to a higher level at about 10 days. The relative risk degree reached a peak at about 15 days, and the ambient temperature had a relatively high relative risk in 0 days and 1 days. The interaction of inhaled particulate matter under 200 g/m3 was more obvious. The long-term trend of the change of death index was related to the temperature and relative humidity. On this basis, the pressure was the influence of death. The main factor of short-term volatility.
The environmental temperature threshold analysis of more than 3. cities showed that the threshold temperature of death risk for all deaths in Beijing, Harbin, Chongqing, Fuzhou and Shantou was 25.63 (+ 0.809), 23.24 (+ 1.114), 29.29 (+ 2.886), 36.06 (+ 0.281) and 31.28 (+ 1.016), when the ambient temperature of each city was higher than the corresponding threshold temperature, respectively. The percentage of increased deaths at 1 degrees centigrade (95% confidence intervals) was 0.99 (0.46~1.52), 1.12 (0.24~2.00), 0.17 (-0.34~0.68), 21.70 (11.99~32.26) and 2.79 (1.05~4.57). The results of Fuzhou were significantly higher, which may be due to data quality questions, so the results need to be further demonstrated.
4. the prediction factors of all urban death factors determined by model fitting analysis of temperature factors are Beijing lag0+lag1+lag24, Chongqing city lag0, Fuzhou city lag0+lag1521, Shantou city lag0+lag24+lag814, Harbin city lag0+lag1+lag814+lag1521, respectively, after the relative humidity factors are included, the prediction factors of each city are Beijing. City lag0+lag1+lag24+lag1521, Chongqing no, Fuzhou no, Shantou city lag0+lag814, Harbin lag0+lag1+lag814+lag1521.
5. the preliminary prediction analysis shows that the influence of high temperature on the death index of residents is mainly concentrated on the lag of 0 to 4 days. The distribution lag nonlinear model can estimate the relative risk degree of death index through the change of temperature and time lag, and then determine the risk of Death Index in the period of desire. Beijing and Chongqing City The corresponding maximum relative risk values are 1.126 and 1.118 respectively during the high temperature period. The corresponding temperature is not the highest daily average temperature. This phenomenon is mainly due to the population's tolerance to high temperature and the corresponding measures to cope with extreme high temperature, and the relative risk of extreme high temperature is reduced; Shantou City, Fuzhou City The maximum relative hazards were 1.310, 1.269 and 1.254, respectively, corresponding to the highest daily mean temperature in each city.
conclusion
1. the lag effect of high temperature is mainly lagging behind 0~4 days, which is an acute effect. There is a long lag effect in the influence of air temperature. The long-term trend of the resident death index is related to the temperature and relative humidity, and the pressure is on the basis of the influence of the temperature and relative humidity, which mainly affects the short-term fluctuation of the death index.
2. only when the temperature factor and the hybrid factor are fitted with the generalized additive model, the effect of the 0 days' temperature in different cities has statistical significance, while the lag time of the temperature effect has a certain difference. When the relative humidity factors are incorporated into the model, the explicit factors of Chongqing and Fuzhou have great changes, indicating the relative humidity factors. There is an obvious mixed effect on the health effects of temperature in Chongqing and Fuzhou.
The 3. distribution lag nonlinear model can estimate the relative risk of the death index of different temperature and different lag time. By observing the temperature change of the day relative to the reference day and determining a lag time, the relative risk of the observation day relative to the reference day is understood, so that the death of the observation day is dead. The relative risk of the index is predicted, and the model can be used to predict the effect of high temperature on health, but the quality of the data has a great influence on the estimation effect of the model.
【學位授予單位】：汕頭大學
【學位級別】：碩士
【學位授予年份】：2011
【分類號】：R188

【參考文獻】

相關期刊論文 前2條

1 劉志剛,陳思靜,錢妙芬;氣候因素與人體疾病研究現狀與展望[J];成都氣象學院學報;1998年01期

2 鄭有飛;氣象與人類健康及其研究[J];氣象科學;1999年04期

，

本文編號：2166118