Active and Passive Microwave Observations to Improve Soil Moisture Predictions Under

Dynamic Vegetation Conditions

J. Judge*, A. Monsivais-Huertero**, K. Nagarajan*, P. W. Liu*

*Center for Remote Sensing, U. of Florida**ESIME Ticoman, Instituto Politecnico Nacional, Mexico

Financial Support from NASA-NIP, NSF-EAR, NASA-THP

UFUNIVERSITY of

FLORIDA

Outline

• Introduction• MicroWEXs• Forward models

– Passive– Active

• Assimilation algorithms• Results• Summary

UF/IFAS

• Application: Assimilate remotely sensed microwave observations to improve root-zone soil moisture estimates in Soil-Vegetation-Atmosphere Transfer (SVAT) models for dynamic vegetation

UF/IFAS

Introduction

• Approach: Couple SVAT + Vegetation growth Microwave models Develop / validate assimilation algorithms for these integrated models

• Problem: Very few detailed (diurnal, season long, high temporal frequency) datasets exist that allow development and testing of coupled models and assimilation algorithms using microwave observations during dynamic vegetation

Microwave water and energy balance experiments (MicroWEXs)

UF/IFAS

• Series of season-long field experiments conducted on a 9-acre field in North Central Florida

• Corn (78 days) – 5 seasons, cotton (130 days) – 2 seasons, Elephant grass (10 months !) – 2 seasons

• Soils - fine sand; heavily irrigated crop

• Observations microwave passive (C, L-band), active (L-

band) soil moisture & temperatures at 2, 4, 8, 16, 32,

64, 120, 170cm soil heat fluxes & physical properties vegetation properties: growth, development,

geometric micro-met, latent heat, sensible heat fluxes,

up/down solar & longwave

MicroWEXs contd.

X X X X

C L

XX

Early MicroWEXs

• RF electronics and antenna: Roger DeRoo & Ruzbeh Akbar @ U. Michigan

• Mechanical Controls and Data Aquistion: UF

• Provide diurnal observations w/ high temporal frequency

UF – CRS L band Automated Scatterometer System MicroWEX contd.

MicroWEX-10

UFLMR MOSS

TMRS

UF/IFAS

42 m X 21m

C L UF- Active UM -Active

75 m X 75 m

UM -Passive

9 m X 9 m

MicroWEX – 10: June – Sept, 2011

10 m X 10 m

MicroWEX contd.

• NASA-THP project Co-I Roger DeRoo, Mahta Moghaddam, Tony England

• Observations in Sweet Corn and Elephant grass• Active & Passive observations under same micromet

conditions

UF/IFAS

Forward Models - Passive• Bare soil: Current brightness models jprovide unrealistic TB during

and immediately following precipitation/irrigation events – challenge for applicability in irrigated agricultural regions…….

VSM0-5 from MicroWEX-5 Soil porosity = 0.37 Rms height = 0.62 cm Correlation length = 8.4 cm

Liu, DeRoo, England, Judge, 2011

UF/IFAS

• Obtain surface roughness, porosity, and VSM in top 1 mm from C-band

Example: rms height =0.41cm, corr. length=8.4cm, porosity 0.55

Passive – Bare Soil contd.

• Dynamic vegetation: opacity formulation dependent upon the growth of the corn crop; compared with the Jackson bW model (Casanova & Judge ( 2008)

Passive – Vegetation

/

2

1exp

)/()(

/700 )(

)(

)(1)(

))(Im(2

2

3

0

0

1987)(Ulaby model dispersion-Dual

e

dhzc

h

B

hzbah

Bz

mkgz

zv

n

zvnzn

dzznk

c

c

e

cc

v

solidsolid

wg

wg

wgwgt

h

t

Liu and Judge, 2011

Passive – Vegetation

Passive – Vegetation

• Combined Wigneron et al. (2007) and Casanova & Judge (2007) : tau is dependent upon angle and polarization

Liu & Judge, 2011UF/IFAS

UF/IFAS

Forward models - Active

• Growing corn: Compared the incoherent and coherent formulations; impact of row structure and location of leaves and ears

Monsivais-Huertero and Judge 2011

Assimilation Algorithms

• Used EnKF-based assimilation for simultaneous estimation of States and Parameters using TB

• Bare soil:

Monsivais-Huertero, Nagarajan, Judge, 2011UF/IFAS

Summary

• MicroWEXs offer season-long high temporal frequency datasets to develop/validate models and assimilation algorithms …. Data available for community use.• Coupled SVAT-Crop models MB and Backscattering

• MicroWEX-10 being conducted during summer 2011 will offer unprecedented diurnal A/P observations, with high temporal frequency

• Current microwave algorithms provide unrealistic brightness during and immediately following the ppt/irrigation events; need a better, more physically-based canopy opacity model during dynamic vegetation

• Looking forward to the diurnal Active & Passive observations during MicroWEX-10 for future improvements in models and assimilation algorithms

UF/IFAS

MicroWEX–8 in 2009

UF/IFAS

• A risk-reduction experiment for the NASA-THP project Co-I Roger DeRoo, Mahta Moghaddam, Tony England - U. Michigan

• Conducted during the corn season; June –August, 2009• Intensive Observation Period (IOP) in August, 2009

Fully mature corn canopy; cleared the footprint bare soil Concurrent active and passive, along w/ Lidar observations Active – U. of Michigan (Moghaddam) Passive – U. of Florida (Judge) Lidar – NCALM; U. of Florida U. of Houston (Shrestha)

MicroWEXs contd.