A Coherent Bistatic Vegetation Model for SoOp Land Applications: Preliminary Simulation Results

Transcription

1 A Coherent Bistatic Vegetation Model for SoOp Land Applications: Preliminary Simulation Results Mehmet Kurum (1), Manohar Deshpande (2), Alicia T. Joseph (2), Peggy E. O Neill (2), Roger H. Lang (3), Orhan Eroglu (1) (1) Mississippi State University, Mississippi State, MS, (2) NASA Goddard Space Flight Center, Greenbelt, MD (3) The George Washington University, Washington, DC, May 25, 2017

2 Outline Motivation Background Formulation Simulation Settings Preliminary Results Future developments Final Remarks 2

3 Signal Spectrum of Interest Frequency Band Ku X C S L P Frequency [GHz] Wavelength [cm] Foliage Penetration Subsurface Imaging Biomass Estimation Agriculture Ocean Ice Snow monitoring 3

4 Opportunistic Signals Sources Frequency Band Ku X C S L P Frequency [GHz] Wavelength [cm] Multiple Scattering Single Scattering Examples Global Navigation Satellite Systems (GNSS) Mobile Satellite Services (MSS) DoD s Mobile Users Objective System (MUOS) DoD s The Ultra High Frequency Follow On (UFO) XM Radio Direct TV And many more! The Question: Can we utilize free illuminators at different frequencies for different Earth land applications? 4

5 Objectives of the Forward Vegetation Model of SoOp 1. Explore new measurement techniques and configuration. 2. Provide intercomparison between measurements from different platforms 3. Deliver test data for training inversion algorithms 4. Broaden the applicability of SoOp to S to P bands for wider utilization of free illuminators. 5. Assess the value of various SoOp signals over land applications 5

6 Linking Models to Measurements REQUIRES: Expertise in knowledge of the inner workings of microwave systems, as well as their relationship to the scientific end products. Vegetation Parameters: Height, Biomass, Type, Health, Etc. Ground Parameters: Surface Soil Moisture Rootzone Soil Moisture Forward Model Inversion Microwave Measurements: SoOp reflectometry System Paremeters (Sensor) Wavelength/Frequency (S, L, and P bands) Polarization (HH, VV, HV, CPs) Incidence angle Resolution Target Paremeters (Vegetation/Ground) Structure (size, orientation, and distribution of scattering surfaces) Surface roughness (relative to wavelength) Dielectric constant (moisture content) Slope angle/orientation 6

7 The Existing Bistatic Vegetation Models Intensity: Incoherent (Radiative Transfer) Models : Bi MIMIC (Linear polarizations) Liang et al., 2005 and ; (Circular/Linear polarizations)wu and Jin, 2014 First order solution Tor Vergata by Ferrazzoli et al., 2010; Multiple Scattering Circular polarization SAVERS by Pierdicca et al., 2014; Uses Tor Vergata Includes antenna effects Complex Voltage: Coherent (Monte Carlo) Models COBISMO by Thirion Lefevre et al., 2010; Single Scattering approximation Linear polarization 7

8 Measurement vs. Theory Incoming Signal: Pencil beam Reflected Signal: Wide beam Most Theoretical Models SmallSats Low Earth Orbit Airplane High Altitude UAS Low Altitude Tower Earth Ground 8

9 Capabilities of New Coherent Bistatic Model Polarimetric effects Fully polarimetric Any combination of linear/circular polarization Interferometric effects Complex Voltage Beamforming Antenna effects Orientation Beamwidth Polarization mixing Beam divergence Sidelobes Cross polarization coupling Configuration effects Altitude Spreading loss over vegetation depth Virtual vegetation Mix vegetation Growing vegetation Seasonal effects 9

10 Beam Divergence X pol pattern R Co pol pattern Variations of the incidence angle across the field of view. Due to the spread of the wave, different particles in the layer experience different incidence directions Vegetation Layer Ground S 10

11 The effect of Beam Divergence on Double Bounce Double Bounce The scattered wave from vertical trunks becomes slightly away from the forward scattering cone. ; ; Weaker scattered fields in the bistatic direction associated with canopy ground interaction. 11

12 Polarization Mixing R,, The direction of polarization vector varies due to the spread of the wave across footprint Vegetation Layer Ground S 12

13 Spreading Loss R Vegetation Layer E 1 E Ground S 13

14 Arbitrary Bistatic Antenna Configuration Total Received Field = Direct + Specular + Diffuse 14

15 Antenna Network Scattering Parameters Total Received Field Direct Specular Reflection Diffuse Scattering The elements of vector (i.e., and ) represent network scattering parameters at the physical antenna ports such as port 1 (along axis) and port 2 (along axis) when the antenna is in reception mode. They are complex voltage quantities and Their magnitude squares are equal to received power at each port. 15

16 Representation of a Vegetated Landscape Discrete Scatterer 16

17 Distorted Born Approximation Independent and single scattering assumption 17

18 Image Technique 18

19 Bistatic Diffuse Scattering Mechanisms Single Bounce Triple Bounce ; ;, ; ;, Double Bounce Double Bounce ; ;, ; ;, 19

20 Contributions in Complex Voltage (Jones vector) Total Received Field Direct... Specular Reflection..,... Diffuse Scattering,..,,... 20

21 Contributions in Power (Modified Stokes Vector) Total Received Power. = Direct... Specular Reflection 4..,.,.,. Diffuse Scattering 4. 1,..,,... 21

22 Antenna Voltage Radiation Pattern Matrix Spherical Basis Linear Polarization Ludwig 3 Basis cos sin sin cos Circular Polarization Basis

23 Antenna Power Radiation Pattern Matrix Four-by-Four Mueller matrix (for Modified Stokes Vector) Re Im Re Im 2Re 2Im 2Re 2Im Re Im Im Re 23

24 Antenna Rotation Rotate about -axis by Rotate about -axis by in elevation cos sin 0 sin cos cos 0 sin sin 0 cos y axis is always parallel to x y ground plane y axis represents H POL while x axis denotes V POL z axis represents boresight direction of the antenna. 24

25 Coordinate Systems T G Reference Coordinate System East North Up (ENU),, R Receive Antenna Coordinate System,, R Transmit Antenna Coordinate System G S T,, Specular Point Coordinate System S,, 25

26 Simulation Configuration 26

27 Vegetated Terrain Parameters Paulownia Trees, Maryland 2006 T1 B1 B2 B3 B4 L1 System Parameters Freq = 400 MHz m SLL = 15 db 50 m XPL = 25 db 100 m EIRP = 0 db 500 m Gr = 0 db L1 B4 L1 B2 B3 B4 B1 T1 B2 sig = 1 ; % rms height in cm sand = 0.80; % percentage of sand clay = 0.07; % percentage of clay rho_b = 1.25 ; % Soil Bulk Density d=2 m d=4 m d=3 m d=4 m 27

28 Specular Contribution (RR and RL) RL RL RL RR RR RR H=20 m H=20 m H=20 m VSM 5% VSM 15% VSM 25% RL RL RL RR RR RR No Free Space Loss No Free Space Loss No Free Space Loss 28

29 Specular Contribution (XX and YY) YY YY YY XX XX XX H=20 m H=20 m H=20 m VSM 5% VSM 15% VSM 25% YY XX YY XX YY XX No Free Space Loss No Free Space Loss No Free Space Loss 29

30 Diffuse Fresnel Zones (RL-pol, theta = 40º) H=20 m H=50 m H=100 m H=500 m 30

31 Diffuse Angular, Height, Circular H=20 m H=50 m H=100 m H=500 m 31

32 Diffuse Angular, Height, Circular, no Ks H=20 m H=50 m No Free Space Loss H=100 m H=500 m 32

33 Diffuse Angular, h =20m, Circular Single Bounce Triple Bounce Double Bounce Double Bounce 33

34 Diffuse Angular, h =100m, Circular Single Bounce Triple Bounce Double Bounce Double Bounce 34

35 Diffuse Angular, h =20m, Circular, no Ks Single Bounce Triple Bounce No Free Space Loss Double Bounce Double Bounce 35

36 Diffuse Angular, h =20m, Linear, no Ks Single Bounce Triple Bounce No Free Space Loss Double Bounce Double Bounce 36

37 Future Developments Include multi layer ground for low frequencies (e.g., P band) to exploit root zone soil moisture estimation Integrate DDM processing for moving transmitter and receiver platforms Develop virtual vegetation for various agricultural and biomes 37

38 Final Remarks Development of advanced models of scattering is fundamental for a FULL exploitation of microwave SoOp signatures! A polarimetric coherent bistatic model has been introduced for vegetated terrains The model successfully generates polarimetric features. Preliminary conclusions at P band: Diffuse term is larger than specular term (10 db or more!) Co and X pol CP responses approaches each other due to depolarization within vegetation Double bounce is dominating. 3 4 db difference is observed over dry and wet soils under ~10 kg/m 2 trees Free space loss may introduce an angular dependence for the reflected signal. It is easy to correct! 38

39 Thank You! Contact: Mehmet Kurum 39