CN105512464A - Method of combining satellite and sites for observation and retrieval of space-time continuous PM2.5 concentration - Google Patents

Abstract

The invention discloses a method of combining a satellite and sites for observation and retrieval of the space-time continuous PM2.5 concentration. Firstly, PM2.5 data of ground sites is subjected to spatial interpolation, and the change rule of each position in time is obtained; then by means of satellite retrieval PM2.5 data in reference time and the change trend of the site interpolation result in the corresponding time period, the estimated value of time under repair is reestablished; finally a regression relation between the un-missing zone of the time under repair and the estimated value of the corresponding position is constructed, the estimated value of the missing zone is corrected based on the regression relation, and the final recovery value is obtained. According to the invention, the time change rule and relation of multi-source data is considered, space-time continuous PM2.5 concentration data can be obtained through reconstruction, and the observation of the ground sites is utilized furthest, The method of combining the satellite and the sites for observation and retrieval of the space-time continuous PM2.5 concentration is high in accuracy and calculating efficiency, easy to implement, strong in extendibility, and high in practical value.

Description

A kind of method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website

Technical field

The invention belongs to remote sensing image processing and Information application field, relate to a kind of method obtaining PM2.5 concentration, be specifically related to a kind of method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website.

Background technology

In the monitoring that air PM2.5 pollutes, the precision of ground station is high, stability strong, but density is low, and coverage is little, cannot carry out comprehensive monitoring and analysis.Remote sensing satellite can provide the gasoloid observation data (AOD) of wide coverage, sets up and after ground station PM2.5 data relationship (AOD-PM2.5), can obtain PM2.5 spatial distribution data by moonscope inverting.But due to the impact of the factors such as cloud, AOD data often exist the situation degrading or lack, thus cause satellite cannot the PM2.5 data of Time Continuous on inverting same position.

Existing method is improve data space coverage to the fundamental purpose that AOD repairs, so that build the inverse model of PM2.5.Its inverting obtains the PM2.5 data of some particular moment, and have ignored discontinuous problem on the time.With the gradual perfection of landing ground PM2.5 monitoring net, website number is continuing to increase, and its data obtained are Time Continuous.Measure and satellite Retrieval data in conjunction with website, consider the Changing Pattern of multi-source data and contact, the high resolving power PM2.5 data obtaining Time Continuous can be rebuild.

Summary of the invention

The object of the invention is to, for the above-mentioned deficiency of prior art, provide a kind of method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website.

The technical solution adopted in the present invention is: a kind of method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website, is characterized in that, comprise the following steps:

Step 1: utilize ground station PM2.5 data to carry out space interpolation;

Step 2: recover region to be restored in satellite Retrieval PM2.5 data, its specific implementation comprises following sub-step:

Step 2.1: for moment t to be repaired _ppM2.5 data, find and there is the t of intact information _mmoment PM2.5 data as a reference;

Step 2.2: utilize t _mthe satellite Retrieval PM2.5 data in moment and ground station interpolation PM2.5 data, and t _pthe ground station interpolation result in moment, estimation t _pthe PM2.5 data in moment;

Step 2.3: for t _pmoment PM2.5 data do not lack district, build the regression relation of the estimated data in satellite Retrieval data and step 2.2; If do not lack too small being not enough in district to build regression relation, then the result in step 2.2 is directly utilized to fill up;

Step 2.4: utilize the linear regression relation in step 2.3, t in aligning step 2.2 _pthe disappearance district PM2.5 estimated value in moment, obtains the final reparation result lacking district.

As preferably, space interpolation described in step 1, employing be learn method of interpolation.

As preferably, in step 2:

Estimation t _pthe PM2.5 data F in moment _pformula be:

$F_p=L_p-\underset{i=1}{\overset{n}{\Σ}}(L_i-F_i);$

Wherein, L _p, L _irepresent t respectively _p, t _ithe ground station interpolation result in moment, n refers to the selected number with reference to the moment, F _irepresent t _imoment satellite Retrieval PM2.5 data;

For t _pmoment PM2.5 data do not lack district, the estimated data F in construction step 2.2 _{p, u}with satellite Retrieval data F _{s, u}regression relation, utilize least-squares estimation to obtain its coefficient for a, b;

To t in step 2.2 _pthe disappearance district PM2.5 estimated value F in moment _{p, o}correct, its expression formula is:

F' _p,o＝a+b·F _p,o；

Wherein, F' _{p, o}represent the estimated value after correcting, the restoration result that namely absent region is final.

The invention has the advantages that:

(1) the PM2.5 data of Time Continuous in each position can be obtained, thus realize seamless monitoring;

(2) the general thinking obtaining Time Continuous PM2.5 data is the reparation of AOD, its can the supplementary of reference less, there is very large uncertainty, present invention effectively prevents this limitation;

(3) after obtaining PM2.5 data based on the inverting of AOD-PM2.5 relation, consider and the contact that multi-source data changes again utilize the time trend of website interpolated data, make it act on maximization.

In a word, the method that the present invention proposes can fill up the vacancy of satellite Retrieval PM2.5 data effectively, obtains restoration result comparatively accurately, realizes large-scale Time Continuous PM2.5 and monitor.

Accompanying drawing explanation

Fig. 1: the process flow diagram of the embodiment of the present invention.

Embodiment

Understand for the ease of those of ordinary skill in the art and implement the present invention, below in conjunction with drawings and Examples, the present invention is described in further detail, should be appreciated that exemplifying embodiment described herein is only for instruction and explanation of the present invention, is not intended to limit the present invention.

The disappearance of satellite data, causes the PM2.5 concentration data of inverting to occur spatial gaps.The satellite Retrieval in the reference moment that data can be utilized intact and ground station interpolation PM2.5 data, and the ground station interpolation result in moment to be repaired, rebuild the PM2.5 data obtaining lacking, thus realize Time Continuous.

Ask for an interview Fig. 1, a kind of method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website provided by the invention, comprises the following steps:

Step 1: utilize ground station PM2.5 data to carry out space interpolation, because website distribution is comparatively sparse, interpolation result is relatively thick, but still maintains temporal variation tendency.By the time trend of thicker interpolation result, as the reference frame of rebuilding high-resolution satellite Retrieval disappearance PM2.5 data.

The space distribution of ground station has different forms, when learning interpolation method (as Ke Lijin, inverse distance-weighting etc.) with utilizing, should consider the distribution characteristics of website.When carrying out space interpolation as utilized Kriging technique, select the kind of variation function according to website distributional pattern feature.

Step 2: recover region to be restored in satellite Retrieval PM2.5 data, its specific implementation comprises following sub-step:

Step 2.1: for moment t to be repaired _ppM2.5 data, find and there is intact information (space all standing as far as possible in principle, as long as in reality and t _pthe disappearance district in moment and do not lack district and all have overlap) t _mmoment PM2.5 data as a reference;

Step 2.2: utilize t _mthe satellite Retrieval PM2.5 data in moment and ground station interpolation PM2.5 data, and t _pthe ground station interpolation result in moment, estimation t _pthe PM2.5 data in moment;

Utilize t _mthe satellite Retrieval PM2.5 data F in moment _mwith ground station interpolation PM2.5 data L _m, and t _pthe ground station interpolation result L in moment _p, estimate moment t to be repaired _phigh resolving power PM2.5 data F _p.Its expression formula is:

F _p＝f(F _m,L _m,L _p)(1)；

Wherein, be the most simply:

$F_p=L_p-\underset{i=1}{\overset{n}{\Σ}}(L_i-F_i)---\left(2\right);$

Wherein, L _p, L _irepresent t respectively _p, t _ithe ground station interpolation result in moment, n refers to the selected number with reference to the moment, F _irepresented for the i-th moment as a reference, that is: can select multiple carry out repairing again with reference to the moment average.In addition, formula (1) relation viewdata rule and make adaptive change.

Step 2.3: for t _pmoment PM2.5 data do not lack district, build the regression relation of the estimated data in satellite Retrieval data and step 2.2; If do not lack too small being not enough in district to build regression relation, then the result in step 2.2 is directly utilized to fill up;

For t _pmoment PM2.5 data do not lack district, the estimated data F in construction step 2.2 _{p, u}with satellite Retrieval data F _{s, u}regression relation, utilize least-squares estimation to obtain its coefficient for a, b.

Step 2.4: utilize the linear regression relation in step 2.3, t in aligning step 2.2 _pthe disappearance district PM2.5 estimated value in moment, obtains the final reparation result lacking district.

To t in step 2.2 _pthe disappearance district PM2.5 estimated value F in moment _{p, o}correct, its expression formula is:

F _p' _,o＝a+b·F _p,o(3)；

Wherein, F' _{p, o}represent the estimated value after correcting, the restoration result that namely absent region is final.

The present invention obtains on PM2.5 data basis at satellite Retrieval, consider rule and the contact of multi-source data time variations, utilize the time trend of ground station interpolated data, with the health data with reference to the moment for reference, the missing information in moment to be repaired is filled up, thus obtains the PM2.5 concentration data of Time Continuous.The method can recover the information lacked more exactly, and has higher counting yield, is easy to drop into practicality.

Should be understood that, the part that this instructions does not elaborate all belongs to prior art.

Should be understood that; the above-mentioned description for preferred embodiment is comparatively detailed; therefore the restriction to scope of patent protection of the present invention can not be thought; those of ordinary skill in the art is under enlightenment of the present invention; do not departing under the ambit that the claims in the present invention protect; can also make and replacing or distortion, all fall within protection scope of the present invention, request protection domain of the present invention should be as the criterion with claims.

Claims (3)

1. observe a method for inverting space and time continuous PM2.5 concentration in conjunction with satellite and website, it is characterized in that, comprise the following steps:

Step 1: utilize ground station PM2.5 data to carry out space interpolation;

Step 2: recover region to be restored in satellite Retrieval PM2.5 data, its specific implementation comprises following sub-step:

Step 2.1: for moment t to be repaired _pdisappearance PM2.5 data, find and there is the t of intact information _mmoment PM2.5 data as a reference;

Step 2.2: utilize t _mthe satellite Retrieval PM2.5 data in moment and ground station interpolation PM2.5 data, and t _pthe ground station interpolation result in moment, estimation t _pthe PM2.5 data in moment;

Step 2.3: for t _pmoment PM2.5 data do not lack district, build the regression relation of the estimated data in satellite Retrieval data and step 2.2; If do not lack too small being not enough in district to build regression relation, then the result in step 2.2 is directly utilized to fill up;

Step 2.4: utilize the linear regression relation in step 2.3, t in aligning step 2.2 _pthe disappearance district PM2.5 estimated value in moment, obtains the final reparation result lacking district.

2. the method in conjunction with satellite and website observation inverting space and time continuous PM2.5 concentration according to claim 1, is characterized in that: space interpolation described in step 1, employing be learn method of interpolation.

3. the method observing inverting space and time continuous PM2.5 concentration in conjunction with satellite and website according to claim 1, is characterized in that, in step 2:

Estimation t _pthe PM2.5 data F in moment _pformula be:

$F_p=L_p-\underset{i=1}{\overset{n}{\Σ}}(L_i-F_i);$

Wherein, L _p, L _irepresent t respectively _p, t _ithe ground station interpolation result in moment, n refers to the selected number with reference to the moment, F _irepresent t _imoment satellite Retrieval PM2.5 data;

For t _pmoment PM2.5 data do not lack district, the estimated data F in construction step 2.2 _{p, u}with satellite Retrieval data F _{s, u}regression relation, utilize least-squares estimation to obtain its coefficient for a, b;

To t in step 2.2 _pthe disappearance district PM2.5 estimated value F in moment _{p, o}correct, its expression formula is:

F′ _p,o＝a+b·F _p,o；

Wherein, F' _{p, o}represent the estimated value after correcting, the restoration result that namely absent region is final.