Soil Moisture Observation Utilizing Reflected GNSS Signals

Transcription

1 Soil Moisture Observation Utilizing Reflected GNSS Signals GNSS-R Tech in Soil Moisture New Data Processing Method Prof. Dongkai YANG Joint African/Asia-Pacific UN-Regional Centers and International Training Workshop on Global Navigation Satellite Systems (GNSS) Aug.

2 GUIDELINE Soil Moisture Observation Utilizing Reflected GNSS Signals I II Ⅲ Ⅳ Ⅴ Introduction Basic Principles Proposed Method Experiment&Results Conclusions

3 Introduction Multipath Effect Signal arrives by More than One Path Transmitter Harmful Effect: Positioning &Timing Error Usually be ELIMINATED Take ADVANTAGES of it! Multipath Receive r

4 Introduction Physical Phenomenon Electromagnetic wave meets reflecting surface Reflection or scattering Reflected Waveform Polarization characteristics Amplitude Phase Soil Moisture Land Surface Material Area Other properties

5 Introduction Research Status -- Experiment In 1993, Martin Neira first proposed PARIS concept, which is an application of ocean altimetry. Since 2000, NASA & University of Colorado made the first experiment to monitor soil moisture based on GNSS-R SMEX In 2009, SMOS launched by ESA can provide a resolution of 50 km space borne observation.

6 Introduction Research Status -- Method In 2010, American professor Larson proposed a new method that signals routinely recorded by GPS receivers installed to measure crustal deformation for geophysical studies could provide a global network of soil moisture sensors, and a new physical model utilizing SNR. The Spanish Starlab used IPT (Interference Pattern Technique). One antenna receives two kinds of signals at the same time, and a notch amplitude, which is sensitive to soil moisture, will appear at the vertical polarization. The conventional method is called ICF (Interferometric Complex Field ). It is widely applied due to its simplicity.

7 Introduction Primary Coverage GNSS-R GlobalNavigationSatelliteSystem Reflection Basic Principles include geometric relationship, signal characteristics, and parameters. Analysis Relationship among parameters to Propose a New data processing Method Experiment for validation by comparison

8 My Contents Soil Moisture Observation Utilizing Reflected GNSS Signals I II Ⅲ Ⅳ Ⅴ Introduction Basic Principles Proposed Method Experiment&Results Conclusions

9 Basic Principles Geometric Model Typical bistatic radar system mode Transmitter: GNSS L-band satellite Reflected signal receiver : LEO(Low Earth Orbit) satellite Not collocated GLONASS GALILEO Receiver GPS Beidou

10 Basic Principles Geometric Model Typical bistatic radar system mode Reflected signal receiver RHCP Antenna Right-Hand Circular Polarized Collects direct GNSS signals LHCP Antenna Left-Hand Circular Polarized Land surface reflected signals RHCP LHCP Direct Signal Reflected Signal

11 Basic Principles Geometric Model Typical bistatic radar system mode Specular reflection: mirrorlike reflection of wave from a surface, in which light from a single incoming direction is reflected into a single outgoing direction. Diffuse reflection: an incident wave is reflected at many angles rather than at just one. RHCP LHCP Direct Signal Reflected Signal Specular Point

12 Basic Principles Geometric Model Typical bistatic radar system mode Specular Direct Signal When diffuse reflection, each ring correspond to a sampling point, or in other word, time delay. RHCP LHCP Reflected Signal Specular Point

13 Basic Principles Geometric Model Typical bistatic radar system mode What do we WANT? Direct Signal Signal Received by LHCP Antenna What can we GET? Incident Signal Reflected Signal RHCP Direct Signal Elevation Angle LHCP Reflected Signal Incident Signal Elevation Angle

14 Basic Principles Geometric Model Typical bistatic radar system mode Because the distance between transmitter and receiver is too far; the distance between receiver and surface is too close RHCP LHCP Direct Signal Elevation Angle Reflected Signal Incident Signal Direct Signal Incident Signal Signal Received by LHCP Antenna Reflected Signal Elevation Angle

15 Basic Principles Signal Characteristic Reflected signal- circular polarization. Right-Hand-Circular-Polarized GNSS signal consists of a vertical and a horizontal polarization component. RHCP LHCP On Reflection Signal would change mostly to LHCP (ele>20 ).

16 Basic Principles Signal Characteristic Reflected signal- circular polarization Therefore,. we use two different antennas to receive different signals separately. RHCP LHCP On Reflection Signal would change mostly to LHCP (ele>20 ).

17 Basic Principles Signal Characteristic Reflected signal - correlation function In reflected signal processing, the correlation function of the local PRN code a and the antenna output signal UR at a time delay of τ is defined as: Self-Correlation Function -Tc In theory Tc -Tc In nature Tc

18 Basic Principles Signal Characteristic Reflected signal - correlation function In theory In nature Interference Attenuation Specular Point Sampling Points Long-tailed-shaped

19 Basic Principles Signal Characteristic Reflected signal - correlation function Interferometric Complex Field PR: reflected correlation power peaks PD: direct correlation power peaks 12 of 34

20 Basic Principles Essential Model Relationship between input & output Output: Soil Moisture Input: Signal Strength Elevation Retrieve Dielectric Constant Reflection Coefficient the square root of reflectivity Reflectivity ICF à Reflectivity

21 Basic Principles Essential Model Relationship between input & output Reflectivity the ratio of reflected signal to incident signal Reflection Coefficient is related to dielectric constant; the square root of reflectivity Reflectivity = ICF = PR/PD

22 My Contents Soil Moisture Observation Utilizing Reflected GNSS Signals I II Ⅲ Ⅳ Ⅴ Introduction Basic Principles Proposed Method Experiment&Results Conclusions

23 Proposed Method Waveform Area New method of data processing Maximum Slope Waveform Peak If we use multiple correlation power values at different time delays The amount of information can be enlarged Its function corresponds to that of a filter.

24 Proposed Method Waveform Area New method of data processing Normalized area with multiple samples that centres around the peak is substituted for the waveform peak of reflected signal. SR(t): Normalized Area Three factors may have effect on the ICF ratio.

25 ❶ Proposed Method Reflectivity Analysis Which reflection coefficient has better performance? All the three kinds of reflectivity could perform well in soil moisture retrieval. To reduce the computation cost Vertical polarization component

26 Proposed Method ❶Reflectivity Analysis Which reflection coefficient has better performance? Substituting R for RRL yield error, but it is negligible (elevation>65 ).

27 ❷ Proposed Method Dielectric Constant Analysis Would elevation angle affect largely on the ICF ratio? ε: the relative complex dielectric constant of the earth surface; ԑ: dielectric constant; ԑ0: dielectric constant of a vacuum; σ: conductivity(mho/m).

28 Proposed Method ❷Dielectric Constant Analysis Would elevation angle affect largely on the ICF ratio? Therefore, dielectric constant relies mainly on the ICF ratio.

29 ❸ Proposed Method Soil Moisture Is calculatedfrom dielectricconstant Volumetric water content is dete rmined by percent by volume S & C: Sand and Clay texture components Constants ß by frequency of incident signal (1.5 GHz) A Semi-Empirical Model Ɛsoil = S C + ( S C) mv + ( S C) mv^2

30 ❸ Proposed Method Soil Moisture Is calculatedfrom dielectricconstant Texture compositions of a soil do not influence significantly on the relationship. Soil moisture depends mainly on the ICF ratio.

31 My Contents Soil Moisture Observation Utilizing Reflected GNSS Signals I II Ⅲ Ⅳ Ⅴ Introduction Basic Principles Proposed Method Experiment&Results Conclusions

32 Experiment & Results Site Description To validate the proposed new method Site Beijing Vegetable Research Center Time On 18 September, 2013 A total of three and a half hours 8:45 a.m. ~ 12:15 a.m. Instrument installation Antenna: height = 2.1m Hardware: GNSS-R receiver Software: data processing (Windows)

33 Experiment & Results Weather Condition To validate the proposed new method The soil was reasonably moist after the precipitation event on the midnight of September 17. During the whole study period, intermittent rainfalls occurred.

34 Experiment & Results TDR Measurement To validate the proposed new method The existing data was collected by TDR (Time-Domain Reflectometer) soil moisture meter every 15 minutes, with probes of 7.5cm and accuracy of ±3.0% volumetric water content.

35 Experiment & Results Data Processing To validate the proposed new method The GNSS-R receiver was designed to choose the satellite with the highest elevation. Elevation angle ranges from 65 to 90 approximately. Satellite Elevation Angle Time Number Elevation 9:10-10: :15-10: :40-11: :43-12:

36 Experiment & Results Data Processing To validate the proposed new method Solid line: GNSS-R measurement is smoothed every 15 minutes Its trend roughly correlates with the trend in the TDR measurement Reflected signal is sensitive to soil moisture.

37 Experiment & Results Data Processing To validate the proposed new method Correlation Analysis The least square method to fit the GNSS-R smoothed data to the existing data Residual sum of squares = 0.01 The two curves in the previous figure show good consistency! GNSS-R Measurement TDR Measurement

38 Experiment & Results Data Processing To validate the proposed new method Smoothing Time It is supposed that the large fluctuation of GNSS-R measurements is caused by stochastic error, because the receiver process signal every second. However, the experiential meteorological data are not recorded with unit of second. They would not change remarkably within several minutes.

39 Experiment & Results Data Processing To validate the proposed new method Smoothing Time To reduce random error, we should first smooth the data before calculating ICF ratio.

40 Experiment & Results Other Results A paper published? Wenjiao Liu, Dongkai Yang, Chaoqun Gao, Qishan Zhang, Soil Moisture Observation Utilizing Reflected Global Navigaiton Satellite System Signals, ICEEICT, 2014.

41 Experiment & Results Other Results A soil moisture retrieval software Demonstrate intermediate parameters, soil moisture, and geometry relationship in real time.

42 My Contents Soil Moisture Observation Utilizing Reflected GNSS Signals I II Ⅲ Ⅳ Ⅴ Introduction Basic Principles Proposed Method Experiment&Results Conclusions

43 Conclusion GNSS-R for soil moisture observation is practical. Reflected signal is sensitive to soil moisture. A new data processing method using reflected signal wave area is valid and has good performance.

44 Conclusion Future Work COMPARISON: proposed & existing method Improve the ACCURACY of the model Soil depth Calibration Vegetation cover Antenna radiation pattern

45 CONTACT US School of Electronic and Information Engineering BEIHANG University Prof. Dongkai YANG Address:, No.37, Xueyuan Rd., Haidian DIstrict, Beijing, CHINA, Phone: (86)

46 THANK YOU