Method for constructing relatively optimal air quality health index

Abstract

A method for constructing a relatively optimal air quality health index. The method comprises: acquiring air pollutants in a region, internal component data of the air pollutants, and health outcome data of the local population; by using a pseudo-Poisson distribution link function, sequentially constructing correlations among the air pollutants, the internal component data of the air pollutants, corresponding daily average concentrations, daily total population mortality, daily outpatient visits and daily inpatient admissions; constructing a multi-pollutant and multi-component Bayesian model on the basis of the air pollutants with the most severely harmful effects during hysteresis, the concentrations of components of the air pollutants, and health outcomes; estimating risk coefficients and weights of multiple pollutants and components, and calculating an excess health risk; establishing a comprehensive air quality health index, and on the basis of different health outcome data of the local population, calculating a plurality of comprehensive air quality health indexes which reflect the health risks associated with the different health outcome data of the local population; and performing comparative analysis on the plurality of comprehensive air quality health indexes, so as to select a relatively optimal air quality health index.

Description

A method for constructing a relatively optimal air quality health index Technical Field

[0001] The present invention relates to the technical field of air quality and health impact early warning and forecasting, and more specifically, to a method for constructing a relatively optimal air quality health index. Background Art

[0002] In recent years, my country has attached increasing importance to the monitoring, early warning and improvement of air pollution levels, emphasizing the comprehensive strengthening of early warning and emergency monitoring construction, deepening forecasting and early warning services, and promoting refined information services for air quality forecasts and early warnings.

[0003] However, as a core indicator of comprehensive air pollution levels in regional forecasting and early warning systems, my country currently primarily uses the Air Quality Index (AQI) to assess air quality. While the AQI takes into account the levels of six commonly monitored pollutants, its calculation method uses a piecewise linear function derived by comparing daily pollutant concentrations with corresponding standard points. Essentially, this reflects the current regional air quality status based on the highest index among the pollutants, reflecting only the risk of the highest pollutant and failing to capture the combined impacts of multiple pollutants on human health. Furthermore, different regions possess unique regional atmospheric complex pollution characteristics, and the AQI calculated based on universal standard points also struggles to capture local pollution patterns and their impact on human health. To address these issues, a Canadian research team pioneered a new index in 2008: the Air Quality Health Index (AQHI). This index, constructed based on the exposure-response relationship between local air pollution and human health, more accurately reflects and assesses air pollution health risks within the context of a specific region and population. However, to date, most studies have constructed the AQHI based on a simple cumulative sum of the excess mortality risks of various pollutants in single-pollutant time series models. While this calculation is relatively simple, it often overestimates the health impact of a single pollutant and, consequently, the combined effects of multiple pollutants. This is primarily due to the following reasons: first, air pollutant concentrations are highly correlated, and the risk coefficient in a single-pollutant model is only the marginal effect value of that pollutant, failing to account for the presence of other pollutants. Second, in reality, people are exposed to a mixture of multiple air pollutants simultaneously, so using a single-pollutant model to estimate health impacts does not truly reflect actual health risks. Third, pollutants, especially particulate matter, have complex internal components, and these models fail to account for their component effects. Finally, using only excess mortality risk as a health outcome to calculate the health index fails to reflect the sensitive health effects of short-term air pollution changes. Therefore, there is an urgent need to design and construct a Composite Air Quality Health Index (C-AQHI) to address these issues. Among them, building a computational model based on multiple pollutants and their components to correctly reflect the exposure-response relationship between multiple pollutants and their components and the impact on human health, as well as how to select the best health outcomes have become the key to constructing a relatively optimal air quality health index.

[0004] Summary of the Invention

[0005] In order to solve the above problems of the shortcomings and defects in the prior art, the present invention provides a method for constructing a relatively optimal air quality health index.

[0006] In order to achieve the above-mentioned purpose of the present invention, the technical solutions adopted are as follows:

[0007] A method for constructing a relatively optimal air quality health index, wherein the method comprises the following steps:

[0008] S1: Obtain data on air pollutants and their components in the area, as well as health outcomes of the local population;

[0009] S2: Using the quasi-likelihood method to adjust for overdispersion in local population health outcome data, a pseudo-Poisson distribution link function was used to sequentially construct data on air pollutants and their constituent components, corresponding to the correlation between daily average concentrations and the daily number of deaths, outpatient visits, and hospitalizations in the total population, i.e., single-pollutant time series model analysis;

[0010] S3: Based on the analysis results of the single-pollutant time series model, a multi-pollutant and component Bayesian model was constructed based on the air pollutant concentration and health outcomes at the lag with the greatest harmful effect;

[0011] S4: Estimate the risk coefficients and weights of multiple pollutants and components based on the Bayesian model of multiple pollutants and components described in S3, and further calculate the excess health risk ER brought by their combined effects. t ;

[0012] S5: Bayesian model based on multiple pollutants and their components and their excess health risk ER t , establish a comprehensive air quality health index C-AQHI,

[0013] S6: Based on the different health outcome data of the local population, calculate multiple comprehensive air quality health indices (C-AQHI) that reflect the health risks of different health outcome data of the local population;

[0014] S7: By comparing and analyzing multiple comprehensive air quality health indices C-AQHI, the relatively optimal air quality health index is screened out.

[0015] Preferably, the air pollutants include particulate matter, SO2, NO2, CO, O3; wherein particulate matter includes PM 2.5 、PM 10 The corresponding internal component data of particulate matter include black carbon, organic matter, sulfate, nitrate, and ammonium salt; the health outcome data of the local population include data on deaths, outpatient visits and hospitalizations and their causes.

[0016] Furthermore, the formula expression of the multi-pollutant and component Bayesian model is as follows:

[0017] Among them, E(Y t ) is the number of deaths Y on day t t, or number of hospitalized patients Y t 、Number of outpatients Y t t represents the tth day; β is the overall effect value of the mixture of air pollutants and their components; X i represents the i-th air pollutant other than particulate matter, where i = 1,…,p, and p represents the total number of air pollutants other than particulate matter; W i represents the coefficient of the i-th air pollutant other than particulate matter; x j represents the jth internal component of a certain particle, where j = 1,…,c, c represents the total number of internal components; W j The coefficient representing the internal component of a certain particulate matter; ε is the residual difference between the internal component of the particulate matter and the particulate matter concentration, which is used to include the impact of residual difficult-to-measure or unmeasured components in the particulate matter on health; W c is the coefficient of the residual of each component; confounders are the factors that affect health caused by various air pollutants and their components; intercept is the possible intercept.

[0018] Furthermore, by assigning the following prior uniform distribution to β in the range of 0 to μ, the best estimate of the overall health effect value of multiple pollutants is obtained.

[0019] Furthermore, by i and W j Specify the following prior Dirichlet distribution so that its cumulative sum is 1, so that W i and W j Equivalent to the weights of each air pollutant and each load component respectively;

[0020] for i=1…p:W i ~Dirichlet(π i );π i ~Uniform(0,1)

[0021] for j=1…c:W j ~Dirichlet(π j );π j ~Unifrom(0,1)

[0022] Among them, Dirichlet() represents the prior Dirichlet distribution; Unifrom represents the prior uniform distribution.

[0023] Furthermore, the formula for calculating the health effects of multiple air pollutants and their components, confounders, is as follows: confounders = ns(time, 7df / year) + ns(hum, 3df) + ns(temp, 4df) + DOW + PH

[0024] Where ns represents natural cubic spline, time represents time trend, df represents degrees of freedom, year represents year, hum represents humidity, temp represents temperature, DOW represents day of the week, and PH represents holiday effect.

[0025] Furthermore, the excess health risk ER t , the formula is as follows:

[0026] Furthermore, a comprehensive air quality health index C-AQHI is established, and its calculation formula is as follows:

[0027] Furthermore, based on the different health outcome data of the local population, a variety of comprehensive air quality health indices (C-AQHI) reflecting the health risks of the local population with different health outcome data are calculated respectively; specifically, based on the number of deaths in the population, a C-AQHI reflecting the excess mortality risk of the population related to short-term air pollution exposure is constructed. M -AQHI, based on the outpatient visit data of the population, reflects the excess morbidity of the population related to short-term air pollution exposure. OV -AQHI, a C-index based on population hospitalization data that reflects the excess risk of illness associated with short-term air pollution exposure. HA -AQHI.

[0028] A computer device includes a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the above method when executing the computer program.

[0029] The beneficial effects of the present invention are as follows:

[0030] Compared to traditional AQI and AQHI models that consider only a single pollutant, or even if they consider multiple pollutants but fail to factor in the inherent component data of air pollutants, the present invention considers a more comprehensive range of influencing factors. Furthermore, a Bayesian model of multiple pollutants and components is constructed using the lagged air pollutant concentrations at the time of greatest harmful effect and the best cause of death outcome. Risk coefficients and weights for these multiple pollutants and components are estimated, and the excess health risk resulting from their combined effects is further calculated. This Bayesian model of multiple pollutants and their components and their excess health risks are then combined to create a comprehensive air quality health index (C-AQHI). The comprehensive air quality health index (C-AQHI) established by the present invention is superior to the AQI established using traditional piecewise linear functions and the AQHI established using traditional generalized additive models. Furthermore, to select the best from the best, the present invention compares and analyzes the comprehensive air quality health index (C-AQHI) corresponding to multiple health outcomes, screening for the optimal air quality health index, ultimately yielding the most sensitive and effective C-AQHI for each health impact, tailored to the region. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] FIG1 is a flow chart of a method for constructing a relatively optimal air quality health index according to the present invention. DETAILED DESCRIPTION

[0032] The following describes the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art will readily appreciate the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the various details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are intended only to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0033] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0034] Example 1

[0035] As shown in FIG1 , a method for constructing a relatively optimal air quality health index includes the following steps:

[0036] S1: Obtain data on air pollutants and their components in the area, as well as health outcomes for the local population. The air pollutants include particulate matter, SO2, NO2, CO, and O3; particulate matter includes PM 2.5 、PM10 ; Particulate matter PM 2.5 The internal component data include black carbon, organic matter, sulfate, nitrate, and ammonium salt; the local population health outcome data include population deaths, outpatient visits and hospitalizations and their causes.

[0037] The air pollutant data in this embodiment mainly comes from the daily monitoring data of the air pollution monitoring station in a certain city from 2016 to 2019, which mainly includes the following six air pollutants: temperature, humidity, PM 2.5 、PM 10 、SO2、NO2、CO、O3. PM 2.5 The component data are derived from the city’s provincial database of daily components from 2016 to 2019.

[0038] Among them, PM 2.5 、PM 10 , SO2, NO2, CO and PM 2.5 The component concentrations are the daily 24-hour average concentrations, O3 is the daily maximum 8-hour average concentration, and these four pollutants are the daily surface concentrations based on the latitude and longitude of individual households. The results are shown in Table 1.

[0039] Table 1 Air pollution and meteorological conditions in a certain city from 2016 to 2019

[0040] The death data of the population comes from the daily death data of the city from 2017 to 2019. The fields mainly include age, gender, time of death, place of death, permanent address before death, underlying cause of death, underlying cause of death ICD 10 th The International Statistical Classification of Diseases and Related Health Problems 10th Revision has a total of 156,000 entries or death data for three years.

[0041] The outpatient and inpatient data were collected from 299 hospitals in the city from 2016 to 2019. The data included patient age, gender, and ICD-10 disease codes. The total number of outpatient visits and hospitalizations over the three years was 18,424,345, and the number of hospitalizations was 4,656,065.

[0042] Table 2 Daily deaths, outpatient visits and hospitalizations in a city

[0043] S2: The quasi-likelihood method is used to adjust for the over-dispersion of local population health outcome data. The pseudo-Poisson distribution link function is used to sequentially construct the data on air pollutants and their components. The correlation between the corresponding daily average concentration and the daily total number of deaths, outpatient visits, and hospitalizations is analyzed, i.e., a single pollutant time series model.

[0044] S3: Based on the analysis results of the single pollutant time series model, in this example, the air pollutant concentration at the lag with the largest harmful effect, that is, the pollutant concentration at lag0, is used to construct a multi-pollutant and component Bayesian model with health outcomes.

[0045] The formula expression of the multi-pollutant and component Bayesian model is as follows:

[0046] Among them, E(Y t ) is the number of deaths Y on day t t , or number of hospitalized patients Y t 、Number of outpatients Y t , t represents the tth day; β is the overall effect value of the mixture of air pollutants and their components, which is determined to be 0.004 through data training in 2016 and 2017 in this example; X i represents the i-th air pollutant other than particulate matter, where i = 1,…,p, and p represents the total number of air pollutants other than particulate matter, which is 4 in this example; W i represents the coefficient of the i-th air pollutant other than particulate matter; x j represents the jth internal component of a certain particle, where j = 1,…,c, c represents the total number of internal components, which is 5 in this example; W j Indicates PM 2.5 The coefficient of the internal load component; ε is the coefficient of the internal load component and PM 2.5 The residual of the concentration is used to include PM 2.5 The health effects of residual unmeasured or unmeasured components in W c is the coefficient of the residual of each component; confounders are the factors that affect health caused by various air pollutants and their components; intercept is the possible intercept.

[0047] In this embodiment, a priori uniform distribution within the range of 0 to μ is specified for β, and data from 2016 and 2017 are used as training data, thereby obtaining the best estimate β of the overall health effect values ​​of multiple pollutants and components as 0.004.

[0048] Specifically, the overall health effect values ​​of multiple pollutants can be optimized for the current data scenario within a reasonable range based on expert knowledge and existing evidence. The formula is as follows: β~Uniform(0,μ)

[0049] Among them, Uniform represents the prior uniform distribution.

[0050] This example also uses W i and W j Specify the following prior Dirichlet distribution so that its cumulative sum is 1, so that W i and W j Equivalent to the weights of each air pollutant and each load component respectively;

[0051] for i=1…p:W i ~Dirichlet(π i );π i ~Uniform(0,1)

[0052] for j=1…c:W j ~Dirichlet(π j );π j ~Uniform(0,1)

[0053] Among them, Dirichlet() represents the prior Dirichlet distribution; Uniform represents the prior uniform distribution.

[0054] In this embodiment, the calculation formula of the factor confounders of the health effects of multiple air pollutants and their components is as follows: confounders = ns(time, 7df / year) + ns(hum, 3df) + ns(temp, 4df) + DOW + PH

[0055] Where ns represents natural cubic spline, time represents time trend, df represents degrees of freedom (7 in this example), year represents year, hum represents humidity, temp represents temperature, DOW represents day of the week, and PH represents holiday effect.

[0056] The calculation formula for confounders is as follows: ns is a natural cubic spline, and the natural fluctuations in the number of deaths / hospitalizations / outpatient visits (such as seasonal effects) are controlled by incorporating the time trend time with 7 degrees of freedom df each year; in addition, cubic spline terms of humidity hum and temperature temp are included to control their nonlinear effects on health outcomes; finally, the day of the week DOW and holiday effects PH are included to eliminate their influence on the association between air pollution and health outcomes.

[0057] S4: Based on the multi-pollutant and component Bayesian model described in S3, estimate the risk coefficients and weights of multiple air pollutants and components in 2018, and further calculate the excess health risk ER brought by their combined effects.t .

[0058] Among them, the excess health risk ER t , the formula is as follows:

[0059] S5: Bayesian model based on multiple pollutants and components and their excess health risk ER t , calculate the comprehensive air quality health index C-AQHI; the calculation formula is as follows:

[0060] S6: Based on the different health outcome data of the local population, calculate multiple comprehensive air quality health indices (C-AQHI) that reflect the health risks of different health outcome data of the local population;

[0061] Specifically, based on the number of deaths in the population, a C coefficient reflecting the excess mortality risk associated with short-term air pollution exposure was constructed. M -AQHI, based on the outpatient visit data of the population, reflects the excess morbidity of the population related to short-term air pollution exposure. OV -AQHI, a C-index based on population hospitalization data that reflects the excess risk of illness associated with short-term air pollution exposure. HA -AQHI.

[0062] In this example, the excess risk estimates of air pollution and component data for different health outcomes and the calculation results of the comprehensive air quality health index are shown in Tables 3 and 4 below.

[0063] Table 3 Excess risk of different health outcomes in the total population associated with different indices per IQR change Note: AQHI is the air quality health index calculated by the traditional single pollutant model; AQI is the air quality index; C-AQHI is the comprehensive air quality health index calculated by the multi-pollutant and component Bayesian model; a Indicates C M -AQHI is a comprehensive air quality health index based on the number of deaths in the population, reflecting the excess mortality risk associated with short-term air pollution exposure; b Indicates C HA -AQHI is a comprehensive air quality health index based on population hospitalization data that reflects the excess risk of illness associated with short-term air pollution exposure; c Indicates C OV -AQHI is a comprehensive air quality health index based on outpatient visit data to reflect the excess morbidity associated with short-term air pollution exposure. Bold indicates a significant association (P < 0.05).

[0064] Table 4 Comparison of different air quality indices from 2016 to 2018 Note: AQHI and C OV -AQHI is calculated based on outpatient data.

[0065] S7: By comparing and analyzing multiple comprehensive air quality health indices C-AQHI, the relatively optimal air quality health index is screened out.

[0066] In this example, a multi-pollutant Bayesian model and its excess health risk ER t , establish a comprehensive air quality health index C-AQHI. Compared with the traditional AQI based on piecewise linear function considering only a single pollutant or the air quality health index AQHI based on generalized additive model considering multiple pollutants, this embodiment considers the multi-pollutant Bayesian model and incorporates component parameters. The AQHI index of each health outcome is relatively optimal under the health outcome. Therefore, the C-AQHI established in this embodiment is M -AQHI, C OV -AQHI, C HA -AQHI's three comprehensive air quality health indices are all optimal.

[0067] For example, as shown in Tables 3 and 4, among the health risk estimates of various health outcomes constructed in this example, the C-AQHI corresponding to the multi-pollutant and component Bayesian model has the highest risk estimate, which is better than the traditional air quality index AQI and the air quality health index AQHI based on the single-pollutant generalized additive model.

[0068] However, in order to select the best from the best, this embodiment establishes C M -AQHI, C OV -AQHI, C HA The three comprehensive air quality health indices (C-AQHI) were compared comprehensively, as shown in Table 3. The C-AQHI (C-AQHI) was obtained by further comparing the three health data (death, hospitalization and outpatient visits). M -AQHI, C OV -AQHI, C HA -AQHI) and determined the C-AQHI (i.e., C OV -AQHI) can best reflect, display and warn of short-term health risks related to changes in air pollution and its components. In this embodiment, we choose to construct a C index based on the outpatient visit data of the population to reflect the excess morbidity of the population related to short-term air pollution exposure. OV -AQHI is the relatively optimal AQHI indicator.

[0069] Example 2

[0070] A computer device includes a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and is characterized in that when the processor executes the computer program, it implements the steps of the method for constructing a relatively optimal air quality health index as described in Example 1.

[0071] The memory and processor are connected using a bus, which can include any number of interconnected buses and bridges. The bus connects various circuits of one or more processors and memories. The bus can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits. These are all well known in the art and are therefore not described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single component or multiple components, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over a wireless medium via an antenna. Furthermore, the antenna receives data and transmits it to the processor.

[0072] In another specific embodiment, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for constructing the relative optimal air quality health index described above are as follows:

[0073] Those skilled in the art will appreciate that all or part of the steps in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a program, which is stored in a storage medium and includes a number of instructions for causing a device (which may be a single-chip microcomputer, chip, etc.) or a processor to execute all or part of the steps in the methods described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0074] Obviously, the above embodiments of the present invention are merely examples for the purpose of illustrating the present invention, and are not intended to limit the embodiments of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A method for constructing a relatively optimal air quality health index, characterized by: The method steps include the following: S1: Obtain data on air pollutants and their components in the area, and health outcomes of the local population; S2: The overdispersion of local population health outcome data was adjusted by quasi-likelihood method, and the data of air pollutants and their components were constructed in sequence with pseudo-Poisson distribution link function, corresponding to the correlation between daily average concentration and the number of deaths, outpatient visits and hospitalizations of the total population per day, i.e. single pollutant time series model analysis; S3: Based on the analysis results of the single pollutant time series model, a multi-pollutant and component Bayesian model was constructed using the air pollutant concentration and health outcomes at the lag with the greatest harmful effect; S4: Based on the multi-pollutant and component Bayesian model described in S3, estimate the risk coefficients and weights of multiple pollutants and components, and further calculate the excess health risk ER caused by their combined effects. t ; S5: Bayesian model based on multiple pollutants and components and their excess health risk ER t , establish a comprehensive air quality health index C-AQHI, S6: Based on the different health outcome data of the local population, calculate a variety of comprehensive air quality health indexes C-AQHI that reflect the health risks of different health outcome data of the local population; S7: By comparing and analyzing various comprehensive air quality health indexes C-AQHI, the relatively optimal air quality health index is screened out.

2. The method for constructing the relative optimal air quality health index according to claim 1, characterized in that: The air pollutants include particulate matter, SO2, NO2, CO, O3; among which particulate matter includes PM 2.5 、PM 10 The corresponding internal component data of particulate matter include black carbon, organic matter, sulfate, nitrate, and ammonium salt; the local population health outcome data include population mortality, outpatient visits and hospitalization and their causes.

3. The method for constructing the relative optimal air quality health index according to claim 2, characterized in that: The formula expression of the multi-pollutant and component Bayesian model is as follows: Among them, E(Y t ) is the number of deaths Y on day t t , or number of hospitalized patients Y t 、Number of outpatients Y t t represents the tth day; β is the overall effect value of the mixture of air pollutants and their components; X i represents the i-th air pollutant other than particulate matter, where i = 1, ..., p, and p represents other air pollutants other than particulate matter The total number of objects; W i represents the coefficient of the i-th air pollutant other than particulate matter; x j represents the jth internal component of a certain particle, where j = 1,…,c, c represents the total number of internal components; W j The coefficient of the internal component of a certain particle; ε is the residual difference between the internal component of the particle and the particle concentration, which is used to include the impact of the residual difficult-to-measure or unmeasured components in the particle on health; W c is the coefficient of the residual of each component; confounders are the factors affecting health caused by various air pollutants and their components; intercept is the possible intercept.

4. The method for constructing the relative optimal air quality health index according to claim 3, characterized in that: By specifying a prior uniform distribution for β in the range of 0 to μ, the best estimate of the overall health effect value of multiple pollutants is obtained.

5. The method for constructing the relative optimal air quality health index according to claim 3, characterized in that: Through W i and W j Specify the following prior Dirichlet distribution so that its cumulative sum is 1, so that W i and W j They are equivalent to the weights of each air pollutant and each load component respectively; for i=1…p:W i ~Dirichlet(π i );p i ~Uniform(0,1) for j=1…c:W j ~Dirichlet(π j );p j ~Uniform(0,1) Among them, Dirichlet() represents the prior Dirichlet distribution; Uniform represents the prior uniform distribution.

6. The method for constructing the relative optimal air quality health index according to claim 3, characterized in that: The calculation formula for the health effects of various air pollutants and their components is as follows: confounders = ns(time, 7df / year) + ns(hum, 3df) + ns(temp, 4df) + DOW + PH Among them, ns represents natural cubic spline, time represents time trend, df represents degree of freedom, year represents year, hum represents humidity, temp represents temperature, DOW represents day of the week; PH represents holiday effect.

7. The method for constructing the relative optimal air quality health index according to claim 3, characterized in that: The excess health risk ER t , the formula is as follows:

8. The method for constructing the relative optimal air quality health index according to claim 3, characterized in that: The comprehensive air quality health index C-AQHI is established, and its calculation formula is as follows:

9. The method for constructing the relative optimal air quality health index according to claim 8, characterized in that: Based on the different health outcome data of the local population, various comprehensive air quality health indices (C-AQHI) reflecting the health risks of different health outcome data of the local population are calculated respectively; specifically, based on the number of deaths among the population, C-AQHI reflecting the excess mortality risk of the population related to short-term air pollution exposure is constructed. M -AQHI, based on the outpatient visit data of the population, reflects the excess morbidity of the population related to short-term air pollution exposure. OV -AQHI, based on population hospitalization data, reflects the excess risk of illness associated with short-term air pollution exposure. HA -AQHI.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

Images