CN106324620A - Tropospheric zenith delay method based not on real-time measurement of surface meteorological data - Google Patents

Abstract

The invention provides a tropospheric zenith delay method based not on real-time measurement of surface meteorological data and having comparative precision as the traditional model. The method utilizes a GPT2 model and bilinear interpolation algorithm to obtain the surface meteorological data of a pre-estimated point according to the latitude and longitude and simplified Julian day of the point, and the obtained meteorological data is input into the Hopfield zenith delay model, and finally the tropospheric zenith delay of the point is obtained. In the process of estimating and predicting the tropospheric zenith delay, the method can get rid of the dependence on the surface meteorological equipment while maintaining the comparable accuracy.

Description

A kind of it is independent of the tropospheric zenith delay method that earth's surface meteorological data is measured in real time

Technical field

The present invention relates to the association area such as satellite navigation and troposphere aeromerric moasurenont, be specifically related to a kind of accurate, disobey Rely the tropospheric zenith delay method measured in real time in earth's surface meteorological data.

Background technology

Along with the development of science and technology, GLONASS (Global Navigation Satellite System, GNSS) be widely applied in military and civilian field, its can provide the user with accurate navigation, location with And the service such as time service.

In GNSS service process, tropospheric inhomogeneities causes the speed of GNSS service electric wave used to slow down and road Footpath bends, and the speed slowed down and the path of bending all cause electric wave to produce delay in time.Prolong in the accurate calculating time During Chi, typically the delay of time is equivalent to the increase in path.Research shows, postpones to be about on tropospheric zenith direction 2.3m, when electromagnetic wave incident angle is reduced to about 10 °, magnitude increases to 20m further.Therefore, tropospheric delay is at GNSS Service process must compensate accordingly.Meanwhile, at target acquisition, the very long baseline interferometry(VLBI (very of folded Clutter in Skywave Radars Long baseline interferometry, VLBI), in the field such as Ground-Based GPS water vapor retrieval and forecast, tropospheric delay is estimated Meter and prediction are also a significant job.

At present, the Main Means calculating tropospheric delay is to utilize mapping function by electric wave prolonging in tropospheric zenith direction Late (hereinafter referred to as zenith delay), angle of incidence direction maps.The numerical value of mapping function normally only with position and electric wave Angle of incidence is correlated with.Therefore, zenith delay is accurately calculated most important in tropospheric delay is estimated and predicted, pertinent literature table Bright, zenith delay model mainly utilizes real time meteorological data set up or be fitted surveying zenith delay numerical value for many years.The former With Hopfield (Hopfield) model and Sa Sitamo Yining (Saastamoinen) model etc. as representative, the latter is with the world Global Satellite troposphere (International Global GPS troposphere, IGGtrop) models etc. are representative.

Wherein, the foundation of the model such as Hopfield and Saastamoinen must obtain in real time surface temperature, air pressure with And the meteorologic parameter such as vapour pressure, therefore model is bigger to the dependence of ground installation.IGGtrop model utilizes cosine function to 5 years Actual measurement zenith delay data are fitted, and zenith delay is expressed as the function of time, height above sea level and longitude and latitude the most at last. The foundation of IGGtrop model is only the matching to conventional data, does not study the Changing Pattern of the factors such as air index.Answering During with, although the mean error in IGGtrop model whole year is relatively low, but estimating that the precision that a certain moment postpones is not good enough, therefore The time resolution rate variance of this model.

It can be seen that all there is corresponding deficiency in above-mentioned zenith delay model: set up based on real-time earth's surface meteorological data Ground installation is relied on bigger by model；And it is relatively low to the model time resolution of measured data matching.Therefore, need badly and seek one Precision and resolution are the highest, and do not rely on the tropospheric zenith delay method of estimation that earth's surface meteorological data is measured in real time.

Summary of the invention

To achieve these goals, the present invention provides one to be independent of earth's surface meteorological data and measures in real time, and time resolution Rate and the highest tropospheric zenith delay method of estimation of precision.

The present invention is independent of the tropospheric zenith delay method that earth's surface meteorological data is measured in real time, including following 2 steps

Step 1: obtain meteorological data

Determine the longitude and latitude of receiver positionλ and height above sea level h₀Parameter；According to longitude and latitude in whole world othermohygrometer 2 Chosen distance receiver in 1 ° × 1 ° grid that (Global Pressure and Temperature 2, GPT2) model provides Four nearest end points, and model parameter A provided by GPT2 model₀、A₁、A₂、B₁、B₂, it is calculated four ends in conjunction with following formula The troposphere parameters such as the temperature T of point, air pressure p, specific humidity Q

$\begin{array}{c}r\left(t\right)=A_0+A_1\cos \left(\frac{doy}{365.25}2\mathrm{\π}\right)+A_2\cos \left(\frac{doy}{365.25}4\mathrm{\π}\right)+\\ B_1\sin \left(\frac{doy}{365.25}2\mathrm{\π}\right)+B_2\sin \left(\frac{doy}{365.25}4\mathrm{\π}\right)\end{array}$

In formula, r (t) represents the meteorologic parameter of t；Doy represents simplification Julian date；When air pressure p and vapour pressure e_w0List When position takes mbar, Q is expressed as:

$Q=\frac{0.622e_{w0}}{p-0.378e_{w0}}$

According to four end points meteorologic parameters T tried to achieve, p and e_w0, in conjunction with following formula, bilinear interpolation goes out the meteorological data of this point ZParameter ZComprise temperature T₀, air pressure P₀, vapour pressure e_w0Information

In formula, Z_0,0、Z_0,1、Z_1,1、Z_1,0Represent the meteorological data of each end points respectively；λ represents longitude and latitude respectively；Parameter Q_0,0、Q_0,1、Q_1,1、Q_1,0Try to achieve with following formula, i.e.

$\left\{\begin{array}{c}{Q}_{00}=(1-p)(1-q)\\ Q_{01}=(1-p)q\\ Q_{10}=p(1-q)\\ Q_{11}=pq\end{array}\right.$

In formula, p=(λ-λ₀₀)/(λ₁₁-λ₀₀)；Parameter lambda therein₀₀、Represent net The longitude and latitude of lower left corner end points in lattice；Parameter lambda₁₁、Represent the longitude and latitude in the upper right corner in grid；

Step 2: the meteorologic parameter of acquisition is input to Hopfield model, is calculated zenith delay, specific as follows:

Surface temperature T that will obtain in step 1₀Bring following formula into and calculate dry, layer heights of roofs h of damp atmosphere_d、h_w

$\left\{\begin{array}{c}{h}_d=40136+148.72(T_0-273.16)\\ h_w=11000\end{array}\right.$

Temperature T at earth's surface that will obtain in step 1₀, air pressure P₀, vapour pressure e_w0Bring following formula into and calculate dry, damp atmosphere Initial refractive index N_d0、N_w0

$\left\{\begin{array}{c}{N}_{d0}=77.6\frac{P_0}{T_0}\\ N_{w0}=3.73\×{10}^5\frac{e_{w0}}{T_0^2}\end{array}\right.$

The N that will obtain in above-mentioned steps_d0、N_w0And h_d、h_w, in conjunction with following formula, calculate the doing of any a height of h in troposphere place, Wet refraction index N_dh、N_wh

$\left\{\begin{array}{c}{N}_{dh}=N_{d0}{\left(\frac{h_d-h}{h_d-h_0}\right)}^4\\ N_{wh}=N_{w0}{\left(\frac{h_w-h}{h_w-h_0}\right)}^4\end{array}\right.$

In formula, h₀Represent local height above sea level；Utilize dry, wet refraction index N_dh、N_whWith dry, the layer heights of roofs of damp atmosphere h_d、h_wDry, the wet stack emission in conjunction with following formula estimation tropospheric zenith

$\left\{\begin{array}{c}{D}_d^z={10}^{-6}\×{\∫}_{h_0}^{h_d}{N}_{dh}dh\\ D_w^z={10}^{-6}\×{\∫}_{h_0}^{h_w}{N}_{wh}dh\end{array}\right.$

According toIt is calculated zenith blind spot D^S。

By above-mentioned calculating process, the present invention can obtain any longitude and latitude zenith delay at any time, and disobeys Rely and measure equipment in Ground Meteorological.

Accompanying drawing explanation

Fig. 1 illustrates that the present invention is independent of the tropospheric zenith delay Method And Principle figure that earth's surface meteorological data is measured in real time；

Fig. 2 illustrates bilinear interpolation schematic diagram；

Fig. 3 illustrates and utilizes GPT2 model bilinear interpolation China part survey station the meteorological data Error Graph of 2012；

Fig. 4 illustrates that the inventive method and Hopfield model calculate the zenith delay error that China's part survey station was in 2012 Comparison diagram；

Fig. 5 shows that the inventive method predicts that certain o'clock is at annual meteorologic parameters in 2016 and zenith delay result figure.

In figure: BJFS is Fangshan, Beijing's survey station；TWTF is peach garden, Taiwan survey station；WUHN is Wuhan, Hubei survey station；XIAN is Xi'an, Shaanxi survey station.

Detailed description of the invention

Below in conjunction with the accompanying drawings, the present invention is discussed in detail is embodied as step.

Fig. 1 illustrates that the present invention is independent of the schematic diagram of the tropospheric zenith delay method that earth's surface meteorological data is measured in real time. From figure 1 it appears that the present invention utilizes bilinear interpolation algorithm to combine GPT2 model obtains the gas needed for Hopfield model As parameter, finally obtain zenith delay according to Hopfield model, including 2 steps.

Step 1: obtain meteorological data

Determine the longitude and latitude of receiver positionλ and height above sea level h₀Etc. parameter；Carry at GPT2 model according to longitude and latitude Four end points that in 1 ° of confession × 1 ° of grid, chosen distance receiver is nearest, and model parameter A provided by GPT2 model₀、 A₁、A₂、B₁、B₂(taking from GPT2 model database), is calculated the temperature T of four end points, air pressure p, specific humidity Q etc. in conjunction with following formula right Fluid layer parameter.

$\begin{array}{c}r\left(t\right)=A_0+A_1\cos \left(\frac{doy}{365.25}2\mathrm{\π}\right)+A_2\cos \left(\frac{doy}{365.25}4\mathrm{\π}\right)+\\ B_1\sin \left(\frac{doy}{365.25}2\mathrm{\π}\right)+B_2\sin \left(\frac{doy}{365.25}4\mathrm{\π}\right)\end{array}$

In formula, r (t) represents the meteorologic parameter of t；Doy represents simplification Julian date.When air pressure p and vapour pressure e_w0List When position takes mbar, Q is represented by:

$Q=\frac{0.622e_{w0}}{p-0.378e_{w0}}$

According to four end points meteorologic parameters T tried to achieve, p and e_w0Meteorological data Z of this point is gone out in conjunction with following formula bilinear interpolationParameter ZComprise temperature T₀, air pressure P₀, vapour pressure e_w0Information.

In formula, Z_0,0、Z_0,1、Z_1,1、Z_1,0Represent the meteorological data of each end points respectively；λ represents longitude and latitude respectively.Parameter Q_0,0、Q_0,1、Q_1,1、Q_1,0Available following formula is tried to achieve, i.e.

$\left\{\begin{array}{c}{Q}_{00}=(1-p)(1-q)\\ Q_{01}=(1-p)q\\ Q_{10}=p(1-q)\\ Q_{11}=pq\end{array}\right.$

In formula, p=(λ-λ₀₀)/(λ₁₁-λ₀₀)；Parameter lambda therein₀₀、Represent net The longitude and latitude of lower left corner end points (such as end points 1 in Fig. 2) in lattice；Parameter lambda₁₁、Represent the upper right corner (such as end points 3 in Fig. 2) in grid Longitude and latitude.

Fig. 3 illustrates and utilizes GPT2 model bilinear interpolation China part survey station meteorological data error of 2012, from Fig. 3 It can be seen that this meteorological data acquisition methods all meets the requirements in the precision of four survey stations of China.

Step 2: the meteorologic parameter of acquisition is input to Hopfield model, is calculated zenith delay.Specific as follows:

Surface temperature T that will obtain in step 1₀Bring following formula into and calculate dry, layer heights of roofs h of damp atmosphere_d、h_w。

$\left\{\begin{array}{c}{h}_d=40136+148.72(T_0-273.16)\\ h_w=11000\end{array}\right.$

Temperature T at earth's surface that will obtain in step 1₀, air pressure P₀, vapour pressure e_w0Bring following formula into and calculate dry, damp atmosphere Initial refractive index N_d0、N_w0。

$\left\{\begin{array}{c}{N}_{d0}=77.6\frac{P_0}{T_0}\\ N_{w0}=3.73\×{10}^5\frac{e_{w0}}{T_0^2}\end{array}\right.$

The N that will obtain in above-mentioned steps_d0、N_w0And h_d、h_wIn conjunction with following formula calculate any a height of h in troposphere place do, wet Refraction index N_dh、N_wh。

$\left\{\begin{array}{c}{N}_{dh}=N_{d0}{\left(\frac{h_d-h}{h_d-h_0}\right)}^4\\ N_{wh}=N_{w0}{\left(\frac{h_w-h}{h_w-h_0}\right)}^4\end{array}\right.$

In formula, h₀Represent local height above sea level.Utilize dry, wet refraction index N_dh、N_whAnd h_d、h_wIt is right to estimate in conjunction with following formula Fluid layer zenith is dry, wet stack emission

$\left\{\begin{array}{c}{D}_d^z={10}^{-6}\×{\∫}_{h_0}^{h_d}{N}_{dh}dh\\ D_w^z={10}^{-6}\×{\∫}_{h_0}^{h_w}{N}_{wh}dh\end{array}\right.$

According toIt is calculated zenith blind spot D^S。

Fig. 4 illustrates that above-mentioned zenith delay computational methods calculate China's part survey station zenith of 2012 with Hopfield model Delay error contrast, figure 4, it is seen that the present invention propose be independent of the real-time measurement model of earth's surface meteorological data with tradition Hopfield model has suitable precision.Therefore, at the GLONASS including Beidou satellite navigation system In, this built-up pattern has a wide range of applications.

One application example of the present invention is as follows:

Utilizing gps satellite that ground subscriber equipment is carried out time service, Timing Receiver is positioned at 0# (34.28 ° of N, 109.03 ° of E), Height above sea level 390m.The zenith delay prediction process of the whole years in 2016 of this situation is as follows:

First, in GPT2 model, carry out trellis search according to the longitude and latitude of 0#, now right during bilinear interpolation Four extreme coordinates answered are respectively 1# (33.5N, 108.5E)；2#(34.5N,108.5E)；3#(34.5N,109.5E)；4# (33.5N,109.5E)；Then, the A provided according to GPT2 model₀、A₁、A₂、B₁、B₂Combine doy etc. parameter and predict that four end points exist Air pressure, temperature and the vapour pressure etc. of 2016, and utilize the meteorologic parameter obtained that 0# is carried out bilinear interpolation；Finally, will The meteorological data that interpolation goes out is input to Hopfield model, it was predicted that annual zenith delay, and Fig. 5 has reacted the meteorological data of prediction And zenith delay result.

Claims (1)

1. it is independent of the tropospheric zenith delay method that earth's surface meteorological data is measured in real time, including following 2 steps

Step 1: obtain meteorological data

Determine the longitude and latitude of receiver positionλ and height above sea level h₀Parameter；According to longitude and latitude in whole world othermohygrometer 2 Four end points that in 1 ° × 1 ° grid that GPT2 model provides, chosen distance receiver is nearest, and the mould provided by GPT2 model Shape parameter A₀、A₁、A₂、B₁、B₂, the troposphere parameters such as the temperature T of four end points, air pressure p, specific humidity Q it are calculated in conjunction with following formula

$\begin{array}{c}r\left(t\right)=A_0+A_1\cos \left(\frac{doy}{365.25}2\mathrm{\π}\right)+A_2\cos \left(\frac{doy}{365.25}4\mathrm{\π}\right)+\\ B_1\sin \left(\frac{doy}{365.25}2\mathrm{\π}\right)+B_2\sin \left(\frac{doy}{365.25}4\mathrm{\π}\right)\end{array}$

In formula, r (t) represents the meteorologic parameter of t；Doy represents simplification Julian date；When air pressure p and vapour pressure e_w0Unit take During mbar, Q is expressed as:

$Q=\frac{0.622e_{w0}}{p-0.378e_{w0}}$

According to four end points meteorologic parameters T tried to achieve, p and e_w0, in conjunction with following formula, bilinear interpolation goes out the meteorological data of this pointParameterComprise temperature T₀, air pressure P₀, vapour pressure e_w0Information

In formula, Z_0,0、Z_0,1、Z_1,1、Z_1,0Represent the meteorological data of each end points respectively；λ represents longitude and latitude respectively；Parameter Q_0,0、 Q_0,1、Q_1,1、Q_1,0Try to achieve with following formula, i.e.

$\left\{\begin{array}{c}{Q}_{00}=(1-p)(1-q)\\ Q_{01}=(1-p)q\\ Q_{10}=p(1-q)\\ Q_{11}=pq\end{array}\right.$

In formula, p=(λ-λ₀₀)/(λ₁₁-λ₀₀)；Parameter lambda therein₀₀、Represent in grid The longitude and latitude of lower left corner end points；Parameter lambda₁₁、Represent the longitude and latitude in the upper right corner in grid；

Step 2: the meteorologic parameter of acquisition is input to Hopfield Hopfield model, is calculated zenith delay, specifically As follows:

Surface temperature T that will obtain in step 1₀Bring following formula into and calculate dry, layer heights of roofs h of damp atmosphere_d、h_w

$\left\{\begin{array}{c}{h}_d=40136+148.72(T_0-273.16)\\ h_w=11000\end{array}\right.$

Temperature T at earth's surface that will obtain in step 1₀, air pressure P₀, vapour pressure e_w0Bring following formula into and calculate dry, the initial folding of damp atmosphere Penetrate index N_d0、N_w0

$\left\{\begin{array}{c}{N}_{d0}=77.6\frac{P_0}{T_0}\\ N_{w0}=3.73\×{10}^5\frac{e_{w0}}{T_0^2}\end{array}\right.$

The N that will obtain in above-mentioned steps_d0、N_w0And h_d、h_w, in conjunction with following formula, calculate dry, the wet folding at any a height of h in troposphere Penetrate index N_dh、N_wh

$\left\{\begin{array}{c}{N}_{dh}=N_{d0}{\left(\frac{h_d-h}{h_d-h_0}\right)}^4\\ N_{wh}=N_{w0}{\left(\frac{h_w-h}{h_w-h_0}\right)}^4\end{array}\right.$

In formula, h₀Represent local height above sea level；Utilize dry, wet refraction index N_dh、N_whWith dry, layer heights of roofs h of damp atmosphere_d、h_w Dry, the wet stack emission in conjunction with following formula estimation tropospheric zenith

$\left\{\begin{array}{c}{D}_d^z={10}^{-6}\×{\∫}_{h_0}^{h_d}{N}_{dh}dh\\ D_w^z={10}^{-6}\×{\∫}_{h_0}^{h_w}{N}_{wh}dh\end{array}\right.$

According toIt is calculated zenith blind spot D^S。