Cover Page for Proposal
Submitted to the
National Aeronautics and
Space Administration
NASA Proposal Number
15-SUSMAP15-0221
NASA PROCEDURE FOR HANDLING PROPOSALS
This proposal shall be used and disclosed for evaluation purposes only, and a copy of this Government notice shall be applied to any reproduction
or abstract thereof. Any authorized restrictive notices that the submitter places on this proposal shall also be strictly complied with. Disclosure of this
proposal for any reason outside the Government evaluation purposes shall be made only to the extent authorized by the Government.
SECTION I - Proposal Information
Principal Investigator
E-mail Address
Phone Number
Brian Hornbuckle
bkh@iastate.edu
515-294-9868
Street Address (1)
Street Address (2)
100 Osborn Drive
3007 Agronomy Hall
City
State / Province
Postal Code
Country Code
Ames
IA
50011-1010
US
Proposal Title :
A SMAP Phenology Product for Croplands
Proposed Start Date
Proposed End Date
Total Budget
Year 1 Budget
Year 2 Budget
Year 3 Budget
06 / 01 / 2016
05 / 31 / 2019
500,911.00
153,983.00
166,618.00
180,310.00
SECTION II - Application Information
NASA Program Announcement Number
NASA Program Announcement Title
NNH15ZDA001N-SUSMAP
Science Utilization of the Soil Moisture Active-Passive Mission
For Consideration By NASA Organization (the soliciting organization, or the organization to which an unsolicited proposal is submitted)
NASA , Headquarters , Science Mission Directorate , Earth Science
Date Submitted
Submission Method
01 / 20 / 2016
Electronic Submission Only
Type of Application
Predecessor Award Number
Grants.gov Application Identifier
Applicant Proposal Identifier
Other Federal Agencies to Which Proposal Has Been Submitted
New
International Participation
Type of International Participation
No
SECTION III - Submitting Organization Information
DUNS Number
CAGE Code
005309844
5J949
Employer Identification Number (EIN or TIN)
Organization Type
2A
Organization Name (Standard/Legal Name)
Company Division
Iowa State University, Ames
Organization DBA Name
Division Number
IOWA STATE UNIVERSITY
Street Address (1)
Street Address (2)
1350 BEARDSHEAR HALL
City
State / Province
AMES
IA
Postal Code
Country Code
50011
USA
SECTION IV - Proposal Point of Contact Information
Name
Email Address
Phone Number
Brian Hornbuckle
bkh@iastate.edu
515-294-9868
SECTION V - Certification and Authorization
Certification of Compliance with Applicable Executive Orders and U.S. Code
By submitting the proposal identified in the Cover Sheet/Proposal Summary in response to this Research Announcement, the Authorizing Official of the proposing organization (or the individual
proposer if there is no proposing organization) as identified below:
•
certifies that the statements made in this proposal are true and complete to the best of his/her knowledge;
•
agrees to accept the obligations to comply with NASA award terms and conditions if an award is made as a result of this proposal; and
•
confirms compliance with all provisions, rules, and stipulations set forth in this solicitation.
Willful provision of false information in this proposal and/or its supporting documents, or in reports required under an ensuing award, is a criminal offense (U.S. Code, Title 18, Section 1001).
Authorized Organizational Representative (AOR) Name
AOR E-mail Address
Phone Number
Andrea Rich
egrants@iastate.edu
515-294-5225
AOR Signature (Must have AOR's original signature. Do not sign "for" AOR.)
FORM NRESS-300 Version 3.0 Apr 09
Date
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION VI - Team Members
Team Member Role
Team Member Name
Contact Phone
E-mail Address
PI
Brian Hornbuckle
515-294-9868
bkh@iastate.edu
Organization/Business Relationship
Cage Code
DUNS#
Iowa State University, Ames
5J949
005309844
International Participation
U.S. Government Agency
Total Funds Requested
No
0.00
Team Member Role
Team Member Name
Contact Phone
E-mail Address
Co-I
Petruta Caragea
515-294-5582
pcaragea@iastate.edu
Organization/Business Relationship
Cage Code
DUNS#
Iowa State University, Ames
5J949
005309844
International Participation
U.S. Government Agency
Total Funds Requested
No
0.00
Team Member Role
Team Member Name
Contact Phone
E-mail Address
Collaborator
Shannon Brown
818-393-0773
Shannon.T.Brown@jpl.nasa.gov
Organization/Business Relationship
Cage Code
DUNS#
Jet Propulsion Laboratory
23835
095633152
International Participation
U.S. Government Agency
Total Funds Requested
No
Other
0.00
Team Member Role
Team Member Name
Contact Phone
E-mail Address
Collaborator
Sidharth Misra
734-846-1071
sidharth.misra@jpl.nasa.gov
Organization/Business Relationship
Cage Code
DUNS#
Jet Propulsion Laboratory
23835
095633152
International Participation
U.S. Government Agency
Total Funds Requested
No
Other
0.00
FORM NRESS-300 Version 3.0 Apr 09
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION VII - Project Summary
While the primary goal of NASA’s Soil Moisture Active Passive (SMAP) mission is to produce global maps of soil moisture, SMAP
is also sensitive to vegetation covering Earth’s surface. SMAP’s Dual–Channel Algorithm (DCA) and Microwave Polarization Ratio
Algorithm (MPRA) use the two polarizations of measured L–band brightness temperature to simultaneously retrieve both soil
moisture and τ, the optical thickness of vegetation. The τ parameter characterizes the degree to which radiation emitted by the soil is
attenuated as it passes through the vegetation canopy on its way to being collected by the satellite. In the case of agricultural crops, the
amount of attenuation will change as crops progress through different phenological stages. We have found that τ retrieved by another
L–band satellite, the European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) mission, reaches a maximum value in the
U.S. Midwest at a specific reproductive stage of corn. Our initial investigation into SMAP τ has determined that it exhibits many of the
same characteristics. However, there are no current plans to utilize τ retrieved by SMAP.
Our long–term goal is to determine how agricultural crops will be affected by climate change. Here, we propose to develop a new
SMAP data product that could be used to track crop phenology, the dates each year at which crops reach certain developmental stages
in response to environmental conditions. We hypothesize that we will be able to leverage what we have learned regarding SMOS τ
and corn phenology to develop this new SMAP product. Our rationale is that this product would contain new information on crop
phenology that can not be detected by current satellite sensors, and hence add significant value to the SMAP mission. We are uniquely
positioned to conduct this research. PI Hornbuckle is an agronomist, an expert in passive remote sensing at L–band, and has reported
on SMOS τ in the peer–reviewed literature. Co–I Caragea is an expert in analyzing data with dependence and developing methodology
related to dynamic linear models that are suited to applications with nonstationary data which display evolving characteristics over
time.
Specifically, we propose to pursue the following objectives.
1. Implement an optimal smoothing method to characterize the SMAP DCA/MPRA τ behavior over time, and quantify the associated
variability/noise.
2. During the growing season in major agricultural regions of the world, detect the timing of local minimum and maximum values of
SMAP DCA/MPRA τ, which correspond to different crop phenological stages or agricultural management such as tillage and harvest,
and continuously update the prediction error of these local extrema as they are approached and passed.
3. Deliver post–season estimates of the timing of local minima and maxima of τ with highest precision and exploit the spatial
relationship among data collected in neighboring pixels to improve the estimates.
4. Illustrate the extra value in our new product in the U.S. Corn Belt by contrasting it with: esti- mates of crop phenology made by the
USDA; and phenology predicted using the accumulation of thermal time calculated from local meteorological stations.
5. Identify agronomic signals in SMAP DCA/MPRA τ, such as crop phenology and agricultural management including tillage and
harvest, in major agricultural regions across the world.
6. Create a 10–year record of crop phenology and the timing of agricultural management by integrating records of SMOS and SMAP
τ.
Up to this point, the value of τ has only been seen in the context of soil moisture retrieval. Our proposal is innovative because we believe
τ itself is an important SMAP product which can be exploited to provide new information.
FORM NRESS-300 Version 3.0 Apr 09
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION VIII - Other Project Information
Proprietary Information
Is proprietary/privileged information included in this application?
Yes
International Collaboration
Does this project involve activities outside the U.S. or partnership with International Collaborators?
No
Principal Investigator
Co-Investigator
Collaborator
Equipment
Facilities
No
No
No
No
No
Explanation :
NASA Civil Servant Project Personnel
Are NASA civil servant personnel participating as team members on this project (include funded and unfunded)?
No
Fiscal Year
Fiscal Year
Fiscal Year
Fiscal Year
Fiscal Year
Fiscal Year
Number of FTEs
Number of FTEs
Number of FTEs
Number of FTEs
Number of FTEs
Number of FTEs
FORM NRESS-300 Version 3.0 Apr 09
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION VIII - Other Project Information
Environmental Impact
Does this project have an actual or potential impact on the environment?
No
Environmental Impact Explanation:
Exemption/EA/EIS Explanation:
FORM NRESS-300 Version 3.0 Apr 09
Has an exemption been authorized or an environmental assessment (EA) or an
environmental impact statement (EIS) been performed?
No
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION VIII - Other Project Information
Historical Site/Object Impact
Does this project have the potential to affect historic, archeological, or traditional cultural sites (such as Native American burial or ceremonial grounds) or historic objects
(such as an historic aircraft or spacecraft)?
No
Explanation:
FORM NRESS-300 Version 3.0 Apr 09
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION IX - Program Specific Data
Question 1 : Short Title:
Answer: A SMAP Phenology Product for Croplands
Question 2 : Type of institution:
Answer: Educational Organization
Question 3 : Will any funding be provided to a federal government organization including NASA Centers, JPL, other Federal agencies,
government laboratories, or Federally Funded Research and Development Centers (FFRDCs)?
Answer: No
Question 4 : Is this Federal government organization a different organization from the proposing (PI) organization?
Answer: N/A
Question 5 : Does this proposal include the use of NASA-provided high end computing?
Answer: No
Question 6 : Research Category:
Answer: 9) Earth System Science applications and decision support
Question 7 : Data Management Plan (Part 1)
Answer:
We will share the validation data generated via this project with others on a site associated with the Iowa Environmental Mesonet similar
to the site where we currently archive SMOS data for the U.S. Midwest (http://mesonet.agron.iastate.edu/smos/). Since netCDF format
has evolved as the de facto standard for weather and climate research, we will make our data available in this format. Data will be made
available after we have published the first refereed paper using the data in question or after the end of the project, whichever comes
earlier. This will ensure that we have sufficient experience with the data to evaluate its validity and completeness.
Question 8 : Data Management Plan (Part 2)
Answer:
No additional characters needed.
Question 9 : Team Members Missing From Cover Page:
FORM NRESS-300 Version 3.0 Apr 09
Answer:
None.
Question 10 : This proposal contains information and/or data that are subject to U.S. export control laws and regulations including Export
Administration Regulations (EAR) and International Traffic in Arms Regulations (ITAR).
Answer: No
Question 11 : I have identified the export-controlled material in this proposal.
Answer: N/A
Question 12 : I acknowledge that the inclusion of such material in this proposal may complicate the government's ability to evaluate the
proposal.
Answer: N/A
Question 13 : Does the proposed work include any involvement with collaborators in China or with Chinese organizations, or does the
proposed work include activities in China?
Answer: No
Question 14 : Are you planning for undergraduate students to be involved in the conduct of the proposed investigation?
Answer: No
Question 15 : If yes, how many different undergraduate students?
Answer: N/A
Question 16 : What is the total number of student-months of involvement for all undergraduate students over the life of the proposed
investigation?
Answer: NA.
Question 17 : Provide the names and current year (1,2,3,4) for any undergraduate students that have already been identified.
Answer:
NA.
Question 18 : Are you planning for graduate students to be involved in the conduct of the proposed investigation?
Answer: Yes
FORM NRESS-300 Version 3.0 Apr 09
Question 19 : If yes, how many different graduate students?
Answer: 2
Question 20 : What is the total number of student-months of involvement for all graduate students over the life of the proposed
investigation?
Answer: 2 graduate students x 6 months per year x 3 years = 36 student-months over the life of the proposal.
Question 21 : Provide the names and current year (1,2,3,4, etc.) for any graduate students that have already been identified.
Answer:
Victoria Walker, Iowa State University, 2. The other graduate student has not yet been identified.
FORM NRESS-300 Version 3.0 Apr 09
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Cumulative Budget
Funds Requested ($)
Budget Cost Category
Year 1 ($)
Year 2 ($)
Year 3 ($)
Total Project ($)
A. Direct Labor - Key Personnel
27,556.00
28,383.00
29,235.00
85,174.00
B. Direct Labor - Other Personnel
59,664.00
61,454.00
63,298.00
184,416.00
2
2
2
6
87,220.00
89,837.00
92,533.00
269,590.00
0.00
0.00
0.00
0.00
1,000.00
5,200.00
11,000.00
17,200.00
1,000.00
5,200.00
5,000.00
11,200.00
0.00
0.00
6,000.00
6,000.00
0.00
0.00
0.00
0.00
Tuition/Fees/Health Insurance
0.00
0.00
0.00
0.00
Stipends
0.00
0.00
0.00
0.00
Travel
0.00
0.00
0.00
0.00
Subsistence
0.00
0.00
0.00
0.00
Other
0.00
0.00
0.00
0.00
Total Number Other Personnel
Total Direct Labor Costs (A+B)
C. Direct Costs - Equipment
D. Direct Costs - Travel
Domestic Travel
Foreign Travel
E. Direct Costs - Participant/Trainee Support Costs
0
Number of Participants/Trainees
21,653.00
23,562.00
24,510.00
69,725.00
Materials and Supplies
0.00
0.00
0.00
0.00
Publication Costs
0.00
1,000.00
1,000.00
2,000.00
Consultant Services
0.00
0.00
0.00
0.00
ADP/Computer Services
0.00
0.00
0.00
0.00
Subawards/Consortium/Contractual Costs
0.00
0.00
0.00
0.00
Equipment or Facility Rental/User Fees
0.00
0.00
0.00
0.00
Alterations and Renovations
0.00
0.00
0.00
0.00
21,653.00
22,562.00
23,510.00
67,725.00
109,873.00
118,599.00
128,043.00
356,515.00
44,110.00
48,019.00
52,267.00
144,396.00
153,983.00
166,618.00
180,310.00
500,911.00
0.00
0.00
0.00
0.00
153,983.00
166,618.00
180,310.00
500,911.00
F. Other Direct Costs
Other
G. Total Direct Costs (A+B+C+D+E+F)
H. Indirect Costs
I. Total Direct and Indirect Costs (G+H)
J. Fee
K. Total Cost (I+J)
Total Cumulative Budget
FORM NRESS-300 Version 3.0 Apr 09
500,911.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2016
Budget Type :
05 / 31 / 2017
Budget Period :
Project
1
A. Direct Labor - Key Personnel
Name
Project Role
Base
Cal. Months Acad. Months
Salary ($)
Funds
Summ.
Requested
Fringe
Months
Salary ($)
Benefits ($)
Requested ($)
Hornbuckle , Brian
PI
10,421.00
1
10,421.00
3,283.00
13,704.00
Caragea , Petruta
CO-I
10,534.00
1
10,534.00
3,318.00
13,852.00
Total Key Personnel Costs
27,556.00
B. Direct Labor - Other Personnel
Number of
Personnel
Project Role
2
Graduate Students
2
Total Number Other Personnel
Cal. Months
12
Acad. Months
Summ. Months
Requested
Fringe
Funds
Salary ($)
Benefits ($)
Requested ($)
52,800.00
6,864.00
59,664.00
Total Other Personnel Costs
59,664.00
Total Direct Labor Costs (Salary, Wages, Fringe Benefits) (A+B)
FORM NRESS-300 Version 3.0 Apr 09
87,220.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
06 / 01 / 2016
End Date :
Budget Type :
05 / 31 / 2017
Project
Budget Period :
1
C. Direct Costs - Equipment
Item No.
Equipment Item Description
Funds Requested ($)
0.00
Total Equipment Costs
D. Direct Costs - Travel
Funds Requested ($)
1,000.00
1. Domestic Travel (Including Canada, Mexico, and U.S. Possessions)
0.00
2. Foreign Travel
Total Travel Costs
1,000.00
E. Direct Costs - Participant/Trainee Support Costs
Funds Requested ($)
1. Tuition/Fees/Health Insurance
0.00
2. Stipends
0.00
3. Travel
0.00
4. Subsistence
0.00
Number of Participants/Trainees:
FORM NRESS-300 Version 3.0 Apr 09
Total Participant/Trainee Support Costs
0.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2016
05 / 31 / 2017
Budget Type :
Budget Period :
Project
1
F. Other Direct Costs
Funds Requested ($)
1. Materials and Supplies
0.00
2. Publication Costs
0.00
3. Consultant Services
0.00
4. ADP/Computer Services
0.00
5. Subawards/Consortium/Contractual Costs
0.00
6. Equipment or Facility Rental/User Fees
0.00
7. Alterations and Renovations
0.00
21,653.00
8 . Other: Tuition
21,653.00
Total Other Direct Costs
G. Total Direct Costs
Funds Requested ($)
109,873.00
Total Direct Costs (A+B+C+D+E+F)
H. Indirect Costs
Indirect Cost Rate (%) Indirect Cost Base ($)
IDC
Cognizant Federal Agency:
50.00
Department of Health and Human Services, Shon
Turner, 214-767-3261
88,220.00
Total Indirect Costs
Funds Requested ($)
44,110.00
44,110.00
I. Direct and Indirect Costs
Funds Requested ($)
Total Direct and Indirect Costs (G+H)
153,983.00
J. Fee
Funds Requested ($)
0.00
Fee
K. Total Cost
Funds Requested ($)
Total Cost with Fee (I+J)
FORM NRESS-300 Version 3.0 Apr 09
153,983.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2017
Budget Type :
05 / 31 / 2018
Budget Period :
Project
2
A. Direct Labor - Key Personnel
Name
Project Role
Base
Cal. Months Acad. Months
Salary ($)
Funds
Summ.
Requested
Fringe
Months
Salary ($)
Benefits ($)
Requested ($)
Caragea , Petruta
CO-I
10,850.00
1
10,850.00
3,418.00
14,268.00
Hornbuckle , Brian
PI
10,734.00
1
10,734.00
3,381.00
14,115.00
Total Key Personnel Costs
28,383.00
B. Direct Labor - Other Personnel
Number of
Personnel
Project Role
2
Graduate Students
2
Total Number Other Personnel
Cal. Months
12
Acad. Months
Summ. Months
Requested
Fringe
Funds
Salary ($)
Benefits ($)
Requested ($)
54,384.00
7,070.00
61,454.00
Total Other Personnel Costs
61,454.00
Total Direct Labor Costs (Salary, Wages, Fringe Benefits) (A+B)
FORM NRESS-300 Version 3.0 Apr 09
89,837.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
06 / 01 / 2017
End Date :
Budget Type :
05 / 31 / 2018
Project
Budget Period :
2
C. Direct Costs - Equipment
Item No.
Equipment Item Description
Funds Requested ($)
0.00
Total Equipment Costs
D. Direct Costs - Travel
Funds Requested ($)
5,200.00
1. Domestic Travel (Including Canada, Mexico, and U.S. Possessions)
0.00
2. Foreign Travel
Total Travel Costs
5,200.00
E. Direct Costs - Participant/Trainee Support Costs
Funds Requested ($)
1. Tuition/Fees/Health Insurance
0.00
2. Stipends
0.00
3. Travel
0.00
4. Subsistence
0.00
Number of Participants/Trainees:
FORM NRESS-300 Version 3.0 Apr 09
Total Participant/Trainee Support Costs
0.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2017
05 / 31 / 2018
Budget Type :
Budget Period :
Project
2
F. Other Direct Costs
Funds Requested ($)
0.00
1. Materials and Supplies
1,000.00
2. Publication Costs
3. Consultant Services
0.00
4. ADP/Computer Services
0.00
5. Subawards/Consortium/Contractual Costs
0.00
6. Equipment or Facility Rental/User Fees
0.00
7. Alterations and Renovations
0.00
22,562.00
8 . Other: Tuition
23,562.00
Total Other Direct Costs
G. Total Direct Costs
Funds Requested ($)
118,599.00
Total Direct Costs (A+B+C+D+E+F)
H. Indirect Costs
Indirect Cost Rate (%) Indirect Cost Base ($)
IDC
Cognizant Federal Agency:
50.00
Department of Health and Human Services, Shon
Turner, 214-767-3261
96,037.00
Total Indirect Costs
Funds Requested ($)
48,019.00
48,019.00
I. Direct and Indirect Costs
Funds Requested ($)
Total Direct and Indirect Costs (G+H)
166,618.00
J. Fee
Funds Requested ($)
0.00
Fee
K. Total Cost
Funds Requested ($)
Total Cost with Fee (I+J)
FORM NRESS-300 Version 3.0 Apr 09
166,618.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2018
Budget Type :
05 / 31 / 2019
Budget Period :
Project
3
A. Direct Labor - Key Personnel
Name
Project Role
Base
Cal. Months Acad. Months
Salary ($)
Funds
Summ.
Requested
Fringe
Months
Salary ($)
Benefits ($)
Requested ($)
Hornbuckle , Brian
PI
11,056.00
1
11,056.00
3,483.00
14,539.00
Caragea , Petruta
CO-I
11,176.00
1
11,176.00
3,520.00
14,696.00
Total Key Personnel Costs
29,235.00
B. Direct Labor - Other Personnel
Number of
Personnel
Project Role
2
Graduate Students
2
Total Number Other Personnel
Cal. Months
12
Acad. Months
Summ. Months
Requested
Fringe
Funds
Salary ($)
Benefits ($)
Requested ($)
56,016.00
7,282.00
63,298.00
Total Other Personnel Costs
63,298.00
Total Direct Labor Costs (Salary, Wages, Fringe Benefits) (A+B)
FORM NRESS-300 Version 3.0 Apr 09
92,533.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
06 / 01 / 2018
End Date :
Budget Type :
05 / 31 / 2019
Project
Budget Period :
3
C. Direct Costs - Equipment
Item No.
Equipment Item Description
Funds Requested ($)
0.00
Total Equipment Costs
D. Direct Costs - Travel
Funds Requested ($)
1. Domestic Travel (Including Canada, Mexico, and U.S. Possessions)
5,000.00
2. Foreign Travel
6,000.00
Total Travel Costs
11,000.00
E. Direct Costs - Participant/Trainee Support Costs
Funds Requested ($)
1. Tuition/Fees/Health Insurance
0.00
2. Stipends
0.00
3. Travel
0.00
4. Subsistence
0.00
Number of Participants/Trainees:
FORM NRESS-300 Version 3.0 Apr 09
Total Participant/Trainee Support Costs
0.00
PI Name : Brian
Hornbuckle
Organization Name : Iowa
NASA Proposal Number
15-SUSMAP15-0221
State University, Ames
Proposal Title : A SMAP Phenology Product for Croplands
SECTION X - Budget
Start Date :
End Date :
06 / 01 / 2018
05 / 31 / 2019
Budget Type :
Budget Period :
Project
3
F. Other Direct Costs
Funds Requested ($)
0.00
1. Materials and Supplies
1,000.00
2. Publication Costs
3. Consultant Services
0.00
4. ADP/Computer Services
0.00
5. Subawards/Consortium/Contractual Costs
0.00
6. Equipment or Facility Rental/User Fees
0.00
7. Alterations and Renovations
0.00
23,510.00
8 . Other: Tuition
24,510.00
Total Other Direct Costs
G. Total Direct Costs
Funds Requested ($)
128,043.00
Total Direct Costs (A+B+C+D+E+F)
H. Indirect Costs
Indirect Cost Rate (%) Indirect Cost Base ($)
IDC
Cognizant Federal Agency:
50.00
Department of Health and Human Services, Shon
Turner, 214-767-3261
104,533.00
Total Indirect Costs
Funds Requested ($)
52,267.00
52,267.00
I. Direct and Indirect Costs
Funds Requested ($)
Total Direct and Indirect Costs (G+H)
180,310.00
J. Fee
Funds Requested ($)
0.00
Fee
K. Total Cost
Funds Requested ($)
Total Cost with Fee (I+J)
FORM NRESS-300 Version 3.0 Apr 09
180,310.00
A SMAP Phenology Product for Croplands
Brian Hornbuckle, Principal Investigator
Department of Agronomy
Iowa State University of Science and Technology
bkh@iastate.edu
Petruţa Caragea, Co–Investigator
Department of Statistics
Iowa State University of Science and Technology
Shannon Brown, Collaborator
NASA Jet Propulsion Laboratory
Sid Misra, Collaborator
NASA Jet Propulsion Laboratory
January 20, 2016
Contents
1 Science / Technical / Management
1.1 Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Vegetation Optical Thickness, τ , at L–band . . . . . . . . . . . . . . . . . . . . . .
1.2.1 SMOS τ in the Corn Belt . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.2 SMAP τ in the Corn Belt . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 A New SMAP Science Application . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.1 Timing of Peak τ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.2 Peak τ and Crop Phenology . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.3 A New SMAP Phenology Product for Croplands . . . . . . . . . . . . . . .
1.4 Significance, Impact, and Why Do This Work Now . . . . . . . . . . . . . . . . . .
1.5 Relevance of Proposed Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.6 Technical Approach and Methodology . . . . . . . . . . . . . . . . . . . . . . . . .
1.6.1 Objectives #1 and #2: Smoothing τ and Quantifying Uncertainty . . . . .
1.6.2 Objective #3: Higher–quality Post–season Data . . . . . . . . . . . . . . .
1.6.3 Objectives #4, #5, and #6: Illustrate Value; Identify New Agronomic Signals; Construct a Long–term Record . . . . . . . . . . . . . . . . . . . . . .
1.7 Plan of Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.8 Data Sharing Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
2
2
3
4
5
6
6
9
10
10
11
11
11
15
. 15
. 16
. 16
2 Investigator Curriculum Vitae
19
3 Current and Pending Support
23
4 Proposal Budget
4.1 Budget Narrative . . . . . . . . . . .
4.1.1 Direct Labor . . . . . . . . .
4.1.2 Direct Costs: Equipment . .
4.1.3 Direct Costs: Travel . . . . .
4.1.4 Other Direct Costs . . . . . .
4.1.5 Indirect Costs . . . . . . . . .
4.2 Proposal Personnel and Work Effort
4.3 Facilities and Equipment . . . . . . .
24
24
24
24
24
25
25
25
26
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 1
Science / Technical / Management
1.1
Executive Summary
While the primary goal of NASA’s Soil Moisture Active Passive (SMAP) mission is to produce global
maps of soil moisture, SMAP is also sensitive to vegetation covering Earth’s surface. SMAP’s
Dual–Channel Algorithm (DCA) and Microwave Polarization Ratio Algorithm (MPRA) use the
two polarizations of measured L–band brightness temperature to simultaneously retrieve both soil
moisture and τ , the optical thickness of vegetation. The τ parameter characterizes the degree
to which radiation emitted by the soil is attenuated as it passes through the vegetation canopy
on its way to being collected by the satellite. In the case of agricultural crops, the amount of
attenuation will change as crops progress through different phenological stages. We have found that
τ retrieved by another L–band satellite, the European Space Agency’s Soil Moisture and Ocean
Salinity (SMOS) mission, reaches a maximum value in the U.S. Midwest at a specific reproductive
stage of corn. Our initial investigation into SMAP τ has determined that it exhibits many of the
same characteristics. However, there are no current plans to utilize τ retrieved by SMAP.
Our long–term goal is to determine how agricultural crops will be affected by climate change.
Here, we propose to develop a new SMAP data product that could be used to track crop phenology,
the dates each year at which crops reach certain developmental stages in response to environmental
conditions. We hypothesize that we will be able to leverage what we have learned regarding SMOS τ
and corn phenology to develop this new SMAP product. Our rationale is that this product would
contain new information on crop phenology that can not be detected by current satellite sensors,
and hence add significant value to the SMAP mission. We are uniquely positioned to conduct this
research. PI Hornbuckle is an agronomist, an expert in passive remote sensing at L–band, and has
reported on SMOS τ in the peer–reviewed literature. Co–I Caragea is an expert in analyzing data
with dependence and developing methodology related to dynamic linear models that are suited to
applications with nonstationary data which display evolving characteristics over time.
Specifically, we propose to pursue the following objectives.
1. Implement an optimal smoothing method to characterize the SMAP DCA/MPRA τ behavior
over time, and quantify the associated variability/noise.
2. During the growing season in major agricultural regions of the world, detect the timing of local
minimum and maximum values of SMAP DCA/MPRA τ , which correspond to different crop
phenological stages or agricultural management such as tillage and harvest, and continuously
update the prediction error of these local extrema as they are approached and passed.
3. Deliver post–season estimates of the timing of local minima and maxima of τ with highest
2
precision and exploit the spatial relationship among data collected in neighboring pixels to
improve the estimates.
4. Illustrate the extra value in our new product in the U.S. Corn Belt by contrasting it with: estimates of crop phenology made by the USDA; and phenology predicted using the accumulation
of thermal time calculated from local meteorological stations.
5. Identify agronomic signals in SMAP DCA/MPRA τ , such as crop phenology and agricultural
management including tillage and harvest, in major agricultural regions across the world.
6. Create a 10–year record of crop phenology and the timing of agricultural management by
integrating records of SMOS and SMAP τ .
Up to this point, the value of τ has only been seen in the context of soil moisture retrieval.
Our proposal is innovative because we believe τ itself is an important SMAP product which can
be exploited to provide new information. We expect the outcomes of this work will be a crop
phenology product that is sensitive to important reproductive stages of crop development that can
not be detected by visible and near–infrared vegetation indices because they occur at a point in
the season after the maximum value of leaf area has accumulated and are therefore “hidden” from
view, when they can only be “seen” at the longer L–band wavelength. In addition, our product will
be sensitive to the timing of agricultural management such as tillage and harvest. Furthermore,
this product will be available at a spatial resolution that is at least 10–times better than USDA
estimates. We believe our work will have impact in four primary areas. First, the detection of crop
stress (which results in delayed or accelerated crop phenology) caused by lack of nutrients, water,
pests, or other environmental effects. Second, forecasting of crop yield. Third, weather forecasting
through a better determination of the time at which crop transpiration will cease. And fourth,
future assessment of changes in crop phenology, agricultural management, and plant biotechnology
caused by, and in reaction to, climate change.
1.2
Vegetation Optical Thickness, τ , at L–band
NASA’s Soil Moisture Active Passive (SMAP) satellite carries a microwave radiometer that operates
at L–band (f = 1.4 GHz, λ = 21 cm). At L–band vegetation is semi–transparent and consequently
Earth’s terrestrial brightness temperature is sensitive to the water content of the first few cm of
the soil surface (e.g. Escorihuela et al., 2010).
While semi–transparent, the confounding influence of vegetation is the single most important
factor affecting the retrieval of soil moisture from measured brightness temperature (e.g. Holmes
et al., 2008). SMAP uses a zero–order solution of radiative transfer commonly called the τ −ω
model to account for the effect of vegetation (e.g. Wigneron et al., 2007).
TB = Tsoil (1 − Rsoil ) e−τ / cos θ
(1.1)
−τ / cos θ
+ (1 − ω) Tveg (1 − e
)
+ (1 − ω) Tveg (1 − e−τ / cos θ ) Rsoil e−τ / cos θ
In (1.1): TB is the brightness temperature; Tsoil is the effective temperature of the soil; Rsoil is
the soil surface reflectivity (and 1 − Rsoil the emissivity of the soil surface); τ is the vegetation
optical thickness; θ is the observation (or incidence) angle; ω is the single–scattering albedo of the
vegetation canopy; and Tveg is the temperature of the vegetation. The three terms in (1.1) represent
the three processes that contribute to the overall TB : emission from the soil that is attenuated as it
3
Figure 1.1: We have a qualitative understanding of τ , the optical thickness of the land surface, at
L–band in the Corn Belt. It increases in response to tillage and increased soil surface roughness in
the early spring, decreases in response to subsequent rainfall, increases in June and July as crops
grow, decreases during crop senescence, and increases again in response to fall tillage. From Patton
and Hornbuckle (2013).
passes through the vegetation; emission from the vegetation itself; and emission from the vegetation
that is scattered by the soil surface and attenuated as it passes back through the vegetation canopy.
The τ parameter represents the degree to which microwave radiation is attenuated by vegetation.
Since absorption leads to emission (Kirchoff’s law), the emission of radiation from the vegetation
itself also depends on τ . It has been shown that τ is directly proportional to the column density
of water stored in vegetation (the mass of water in vegetation tissue per ground area), commonly
called vegetation water or vegetation water content (VWC) (e.g. Jackson and Schmugge, 1991).
τ = b × VWC
(1.2)
The proportionality constant is called the “b–parameter” and is a function of frequency, polarization, and the type of vegetation. It is apparent from (1.2) that τ increases as the amount of
vegetation increases.
1.2.1
SMOS τ in the Corn Belt
Another L–band satellite radiometer, the European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) mission, was launched in late 2009 (Kerr et al., 2010). SMOS observations have a
temporal frequency and spatial resolution similar to SMAP, and SMOS retrieves the τ parameter
by using TB measured at multiple incidence angles θ to optimize (1.1) using a cost function (Kerr
et al., 2012). Patton and Hornbuckle (2013) evaluated SMOS τ in the Corn Belt of the Midwest
4
U.S. Part of their work is summarized in Figure 1.1. In the upper left of Figure 1.1 is the record
of SMOS τ for a pixel that lies in north central Iowa, a state in the Corn Belt. Both the actual
data (which is noisy) and 21–day moving average are shown. Crops in the Corn Belt are normally
planted in April and May. As the crops grow, τ steadily increases as predicted by (1.2) until a
peak value is reached later in the summer. This behavior can be seen in Region III of Figure 1.1.
Annual crops like corn and soybean eventually stop growing and die, in a process called senescence.
During this process they slowly lose their moisture. When the crops are harvested in October or
early November, they contain little water. Hence the slow decline in τ , also as predicted by (1.2),
that occurs over a period of a few months (Region IV of Figure 1.1).
Outside of the growing season, field activities (agricultural management) such as tillage, planting, and the application of fertilizer will roughen the soil surface and increase its microwave emissivity. A “roughness parameter” h modifies Rsoil in (1.1) such that
Rsoil = R e−h
(1.3)
and R is the reflectivity of a specular (flat) soil surface. SMAP and SMOS both assume that soil
surface roughness does not change over time and hence use a constant value of h in their soil
moisture retrieval algorithms. Patton and Hornbuckle (2013) found that, numerically, changes in
soil surface roughness have a similar overall effect as changes in the amount of vegetation on TB
and therefore are interpreted as changes in τ by SMOS:
2
∆τ
(1.4)
cos θ
(for vertically–polarized TB ). Consequently, values of τ retrieved by SMOS outside of the growing
season (Regions I, II, and V in Figure 1.1) are actually caused by changes in soil surface roughness
due to field activities that roughen the soil surface and subsequent precipitation that erodes and
flattens the soil surface (Zobeck and Onstad , 1987).
In the bottom–right of Figure 1.1 is a cartoon illustration of our overall qualitative understanding of SMOS τ in the Corn Belt. In late March and early April, agricultural management
activities such as tillage and planting roughen the soil surface and this is interpreted by SMOS as
an increase in τ . Subsequent rainfall that reaches the soil surface (and not intercepted by crop
leaves) erodes soil roughness and decreases τ . After canopy closure, crop growth dominates and
τ increases dramatically, peaks at some point during the growing season, and then falls as crops
senesce and dry out. Management activities that occur after harvest roughen the soil surface and τ
increases again in October and November, even though there is no growing vegetation. Throughout
the year τ changes in response to other vegetation (e.g., trees and pasture) but these changes are
small compared to those related to crop management in the Corn Belt.
∆h =
1.2.2
SMAP τ in the Corn Belt
The SMAP mission is currently evaluating the performance of four distinct soil moisture retrieval
algorithms. Two of the four, the SMAP Dual–Channel Algorithm (DCA) and the Microwave
Polarization Ratio Algorithm (MPRA), use the two polarizations of L–band TB observed at θ = 40◦
to retrieve both soil moisture and τ . We have analyzed SMAP DCA τ for 2015 and find that it
exhibits nearly the same characteristics as SMOS τ . SMAP DCA τ and SMOS τ for a pixel covering
the South Fork SMAP Core Validation Site in Iowa is shown in Figure 1.2. Note that SMAP τ
is even higher than SMOS τ in Region I after spring management activities. SMAP τ decreases
in Region II in response to rainfall, increases in Region III as crops grow, decreases in Region IV
as crops senesce and are harvested, and increases again (even more than SMOS τ ) in Region V in
response to fall fieldwork.
5
Figure 1.2: Our initial investigation of SMAP DCA τ indicates that it is as sensitive to changes in
soil surface roughness and crop growth in the Corn Belt as SMOS τ .
1.3
A New SMAP Science Application
Note in Figures 1.1 and 1.2 that τ reaches a local maximum sometime between day of year 220
and 240, during the month of August. Immediately before and after this point in the growing
season, the majority of the τ signal is due to the growth and development of crops. Does the peak
value of τ occur at the same time each year?
1.3.1
Timing of Peak τ
Hornbuckle et al. (2016) examined four years of SMOS τ in order to answer this question. What
they found is shown in Figure 1.3. They chose 30 SMOS pixels in Iowa, a state in the Corn Belt, in
which at least 75% and as much as 85% of the land area within each pixel was devoted to growing
corn and soybean. Both of these crops are planted as early as mid–April and as late as early June
(corn followed by soybean), and are harvested in October or early November. About 60% of the
cropped area is corn and 40% soybean.
SMOS τ from 2010 to 2013 was smoothed using a technique called functional data analysis
(FDA) (Ramsay et al., 2013). This smoothing procedure is illustrated in the bottom–left of Figure 1.3. The resulting τ for one of the 30 pixels (pixel 2 in north–central Iowa) is shown in the
bottom–left of Figure 1.3. Note that the peak value of τ did not occur at the same time each year.
Peak τ occurred the earliest in 2012 (near day of year 210) and more than 30 days later in 2013
(near day of year 243). Referring to the cartoon illustration in the bottom–right of Figure 1.1 and
recalling (1.2), the data indicate that the order in which the crops within this pixel reached their
6
Figure 1.3: After smoothing, SMOS τ in the Corn Belt peaks on a different day each year. We
propose to exploit the information contained within this τ signal.
highest combined VWC was earliest in 2012, followed by 2010, 2011, and finally 2013.
Hornbuckle et al. (2016) were able to explain the timing of peak τ as follows. First, plants proceed through different developmental or phenological stages like leaf–out, flowering, and senescence
according to the accumulation of thermal time. If the rate of development is a linear function of
temperature, then
Z t
0
c(t) =
T (t ) − Tbase dt0
(1.5)
0
where: c(t) is the accumulated thermal time from 0 to time t; T (t) is the temperature of the
plant at time t; and Tbase is the base temperature, the temperature at which plant development no
longer can occur (e.g. Campbell and Norman, 1998). For crops, thermal time is often referred to
as growing degree days (GDD), which can be calculated with sufficient accuracy as follows.
GDD =
Thigh + Tlow
− Tbase
2
(1.6)
Here: Thigh and Tlow are the daily high and low air temperature recorded at a nearby meteorological
station; and a unit time period of a day is assumed, such that GDD have units of ◦ C days. For corn,
Tbase ≈ 10 ◦ C, and a certain number of GDD are needed to progress from one phenological stage
to another. For example, emergence (when the corn plant is first visible above the soil surface)
7
Figure 1.4: At left, destructive sampling of corn and soybean indicates the overall VWC (and thus
τ ) of a satellite pixel in the Corn Belt (blue line) will peak at 1000 ◦ C day after planting of corn. At
right, a significant correlation exists between the day at which 1000 ◦ C day has been accumulated
after planting and the day of year of the local maximum of SMOS τ .
requires about 60 ◦ C day after planting, while about 1400 ◦ C day after planting are required for a
corn plant to reach full maturity.
Second, Hornbuckle et al. (2016) found that a typical satellite pixel in Iowa reaches its maximum
value of VWC (and therefore τ according to (1.2)) at about 1000 ◦ C day after planting of corn.
They determined this by destructively sampling the VWC (or equivalently, the vegetation water
column density) of hundreds of corn and soybean plants in several Nebraska fields. These data
are shown on the left–hand side of Figure 1.4. Note that the VWC of corn is much larger than
the VWC of soybean. Within a typical Iowa satellite pixel, 60% of the cropped area will be in
corn and 40% in soybean. When the VWC of corn and soybean is combined to match this 60/40
ratio, the result is the blue line on the left–hand side of Figure 1.4. This blue line, representing the
effective VWC of a hypothetical but typical pixel in Iowa, peaks at 1000 ◦ C day after planting of
corn. Therefore, according to (1.2), τ should also peak at 1000 ◦ C day.
When Hornbuckle et al. (2016) regressed the day of year at which 1000 ◦ C day had been
accumulated after planting (using USDA estimates of when 50% of the corn in the pixel had been
planted) against the day of year at which the maximum value of SMOS τ occurred, they found a
significant one–to–one relationship (R = 0.73, p = 2.6 × 10−21 ). This regression is shown on the
right–hand side of Figure 1.4. In summary,
SMOS τ reaches its peak growing season value 1000 ◦ C day after planting of corn.
Note how all 30 pixels in 2012 reached 1000 ◦ C day GDD and their maximum value of τ earliest
in 2012 and latest in 2013. Temperatures in Iowa in 2012 were unusually warm, and as a result,
crops reached 1000 ◦ C day GDD on a much earlier date than in 2013 when temperatures were lower,
on average, than during any of the other three years. While the relationship shown on the right–
hand side of Figure 1.4 is not perfect, it is impressive considering the nature of the data. USDA
estimates of planting are made by human surveyors in nine separate Crop Reporting Districts which
cover the entire state of Iowa, and the data used to calculate GDD using (1.6) were obtained from
NWS COOP stations nearest to each pixel.
8
Figure 1.5: The day of year at which SMOS τ reaches a local maximum during the growing season
corresponds to the R3 reproductive stage of corn.
1.3.2
Peak τ and Crop Phenology
Finally, Hornbuckle et al. (2016) were able to determine the phenological stage of corn at 1000 ◦ C day
GDD, and hence the phenological stage at which SMOS τ reaches its maximum value. The destructive sampling of corn in Nebraska included the calculation of VWC as well as recording the
phenological stage of the crop based upon visual identification. For most of the destructive samples,
VWC was highest when the corn had reached the third reproductive stage, R3. After emergence,
corn progresses through vegetative stages during which leaves are added to the plant (V1, V2, V3,
etc.). Once a corn plant generates a tassel (VT) it enters the reproductive stages during which the
kernels in the ear are fertilized (R1) and they grow in size. Full maturity, the point at which the
kernels stop accumulating carbon, occurs at R6. The R3 stage lies between R1 and R6 (Abendroth
et al., 2011). These stages are listed, and some are illustrated, in Figure 1.5.
Hornbuckle et al. (2016) again used USDA survey data to determine the correlation between:
the day of year 50% of the corn in five representative pixels had reached the R3 stage; and the day
of maximum SMOS τ . They found a significant positive correlation (R = 0.81, p = 1.5 × 10−5 ).
These data are shown at the bottom–right of Figure 1.5. In summary,
SMOS τ reaches its peak growing season value at the R3 corn phenological stage.
9
1.3.3
A New SMAP Phenology Product for Croplands
We propose to develop a new SMAP data product from existing SMAP DCA/MPRA τ that can
be used to track the phenology of crops and agricultural management such as tillage and harvest.
This product will consist of:
• an optimally smoothed time series of τ (and associated uncertainty); and
• the day of year at which local maxima and minima are reached (and associated uncertainty).
We hypothesize that we will be able to leverage what we have learned regarding the relationship
between SMOS τ and corn phenology to develop this new SMAP product. We will construct this
new product for major agricultural regions of the world. We will illustrate the extra value in our
new product in the U.S. Corn Belt by contrasting it with: estimates of crop phenology made by
the USDA; and phenology predicted using the accumulation of thermal time calculated from data
collected at local meteorological stations. We will also search for other crop phenological signals, as
well as other agronomic signals such as the timing of tillage and harvest, in the major agricultural
regions of the world.
1.4
Significance, Impact, and Why Do This Work Now
Our proposal is innovative because we propose to use τ , a SMAP product originally thought to
be of value only in the context of soil moisture retrieval, to provide new information regarding
crop phenology and the timing of agricultural management. This information can not be provided
by higher–frequency satellite sensors. Crop reproductive development occurs within the canopy,
“hidden” from the view of visible and near–infrared sensors and microwave radiometers that operate
at shorter wavelengths. However, at L–band crop canopies are semi–transparent and reproductive
development, such as the growth of the ear of a corn plant, can be “seen” at the longer L–band
wavelength. In addition, this information would be available at a spatial resolution that is at least
10–times better than USDA estimates (note that approximately ten satellite pixels are able to fit
within each of the nine Crop Reporting Districts in Iowa shown in Figure 1.5) and in near–real–
time. Furthermore, little information on tillage, a type of agricultural management that may be
detected by SMAP τ , is publicly available. Tillage plays an important role in the partitioning of
rainfall into infiltration and runoff, soil erosion, and biogeochemical cycles. Thus our work would
add significant value to the SMAP mission by addressing “Agricultural Productivity,” one of five
“SMAP Applications Areas.”
Changes in vegetation status have been monitored at higher microwave frequencies (Jones et al.,
2011), but Patton and Hornbuckle (2013) and Hornbuckle et al. (2016) are the first to relate L–band
τ to agricultural management and a specific crop phenological stage. L–band τ can also be more
rigorously justified from a theoretical standpoint than τ produced using observations from higher
frequency satellite radiometers. (1.1) is a zero–order solution of radiative transfer in vegetation
which is only valid when scattering within the canopy is small. Scattering depends on the electrical
size of canopy components (stems, leaves, reproductive organs), the size of these components relative
to the wavelength of radiation. Scattering becomes more significant as the electrical size of potential
scatterers increases. At L–band λ = 21 cm, which is about a factor of 10 larger than the width of
stems and leaves of crops like corn and soybean. On the other hand, the wavelength used by Jones
et al. (2011) at 18.7 GHz is λ = 1.60 cm. Hence (1.1) and thus the SMAP DCA/MPRA is much
more likely to be a true representation of τ than at other microwave frequencies currently available
from existing satellite radiometers.
10
We believe our work will have impact in four primary areas. First, the detection of crop
stress caused by lack of nutrients, water, pests, or extreme heat. Mild stress results in delayed
crop development, while severe stress accelerates crop development. Warmer temperatures and
more variable rainfall in future climates will likely lead to greater crop stress in the future and
impact crop productivity (Lobell et al., 2013, 2014). Second, forecasting of crop yield. Crop
models (e.g. Archontoulis et al., 2014) predict crop development according to the accumulation
of GDD. However, assimilation of crop phenology information, especially during the reproductive
developmental stage, could improve predictions of how weather conditions will impact crop yield.
Third, weather forecasts. Land surface models within weather and climate models control how
energy and water move between Earth’s surface and atmosphere. Some weather and climate models
have incorporated crop models into their land surface model in order to better represent phenology
and its impact on land surface albedo and transpiration (e.g. Levis et al., 2012). Senescence
and the termination of crops and thus transpiration is triggered by the accumulation of GDD.
Assimilation of crop phenology, especially during the reproductive development stages which occur
shortly before senescence, may allow the crop model to better estimate the timing of senescence
and consequently land surface water and energy fluxes. And fourth, future assessment of changes in
crop phenology, agricultural management, and plant biotechnology (new crop hybrids with unique
phenology) caused by, and in reaction to, climate change.
Why do this work now? Data collected by SMAP, SMOS, and future satellite L–band radiometers could be used to create long–term records of crop phenology and the timing of agricultural
management. It is imperative that this research commence while SMOS is most likely to still be
functioning so multiple years of comparisons can be made between SMOS and SMAP τ . SMOS
was launched in late 2009 and was designed to be a three year mission. However, it has recently
been approved to continue operating through the next few years.
1.5
Relevance of Proposed Work
This proposal is responsive to ROSES 2015 A.22 “Science Utilization of the Soil Moisture Active
Passive Mission” because it addresses two of the four types of proposals solicited. First, we propose
to utilize a SMAP product, namely τ from the DCA and/or MPRA, to improve our understanding
of crop phenology and the timing of agricultural management such as tillage, fertilization, and
harvest. Such agricultural activities are directly tied to the global water, energy, and carbon
cycles. Hence our work is relevant to “2.1 Utilization of SMAP Products for Process Studies.”
Second, we propose to make available a new data product, the timing of local maxima and minima
of SMAP DCA and/or MPRA τ . We will use new statistical techniques developed by Co–I Caragea
to identify these local maxima and minima. The values of τ that we use for this new product may
be those distributed by the SMAP mission, or may be our own τ retrievals made using brightness
temperatures of higher spatial resolution provided to us by Collaborators Brown and Misra. Hence
we believe our work is also relevant to “2.3.1 New Algorithms, Novel Topics.”
1.6
1.6.1
Technical Approach and Methodology
Objectives #1 and #2: Smoothing τ and Quantifying Uncertainty
As seen in Figures 1.1, 1.2, and 1.3, SMOS and SMAP τ exhibit high–frequency noise (large
variations on the order of a day). Variations of more than 30% in the value of τ within the month
of August can be seen in Figure 1.2. By this time of the growing season crops in Iowa have reached
11
the reproductive stages during which plants are gradually increasing their mass. On the other hand,
there are diurnal changes in the water content of crops that may be significant at the satellite scale.
The water potential of plant tissue, and consequently plant water content, changes over the course
of a day as a result of transpiration, the movement of water from the soil, into plant roots, through
a plant’s vascular system, and eventually out of the stomata in its leaves (Slatyer , 1967). Hunt et al.
(2011) observed this diurnal change through the analysis of cellular signals propagating through
a field of corn. They found that signal strength was inversely proportional to vegetation water
content. A clear diurnal pattern, with vegetation water content being largest at night and lowest
during daylight hours, appeared when the data was detrended to account for the seasonal change.
While significant diurnal variations in τ at the satellite scale may exist, we propose to simply
identify local maxima and minima, and therefore the use of a smoothing method that removes
natural and possibly artificial high–frequency variations in τ that occur over time periods of less
than a week is appropriate for our work.
Dynamic Linear Models (DLMs)
We plan to use dynamic linear models (DLMs) to smooth SMAP τ . DLMs, also referred to as
state space models, are a large, flexible class of time series models. A distinguishing feature which
renders them superior to traditional time series models is their ability to account for nonstationarity
in the data. Here nonstationarity can be of the first (i.e. non–constant mean) or second order (i.e.,
non–constant variance). The latter type of nonstationarity is usually the more serious challenge
for the traditional time series set–up, but it can be handled elegantly by DLMs when allowing for
time–varying coefficients and variance parameters.
In general, a DLM assumes that a time series (Yt )t≥1 is an observed realization of an underlying
latent process on a p–dimensional state vector (θt )t≥1 , subject to Gaussian random noise. More
specifically, the model is typically defined using the following two equations:
Yt = Ft θt + vt
vt ∼ N (0, Vt )
(1.7)
θt = Gt θt−1 + wt
wt ∼ N (0, Wt )
(1.8)
where Ft is a matrix of known constants, such as covariates, and (vt )t≥1 is the observation error assumed to be a sequence of independent, mean zero Gaussian random variable with variance
(Vt )t≥1 . Gt is a matrix of known constants (also known as the evolution, or transition matrix),
since it defines how the state evolves from time t − 1 to t. (wt )t≥1 is assumed to be a sequence
of independent, mean zero Gaussian random variables with variance (Wt )t≥1 and is assumed independent of (vt )t≥1 . Usually, Ft and Gt are completely specified. This general form of the model
can be modified by imposing constraints on the various model components. For example, setting
(Vt )t≥1 = V implies the variance of observed values is constant over time, and likewise for the
system variances (Wt )t≥1 = W . Setting Gt = I and Wt = 0 reduces the DLM to the static linear
model Yt = Ft θ + vt . Equation (1.7) is referred to as the observation equation, while Equation (1.8)
is often referred to as the state, system, or evolution equation (West and Harrison, 1997).
The DLM specification results in two key conditional independence probabilities. The current
state of the process θt depends only on the previous state θt−1 and is conditionally independent of
the state at all previous time points. Additionally, the observed process at time t, Yt , depends only
on the current state, θt and is conditionally independent of all other observations and states, given
θt . These conditional probability specifications lend themselves very well to recursive estimation
procedures that fall naturally into the Bayesian framework. Therefore, to complete the above
specification of the DLMs, we set a prior on the initial state, call it θ0 , assume it is independent of
12
(vt )t≥1 and (wt )t≥1 , and such that θ0 ∼ N (m0 , C0 ), with m0 and C0 to be specified as part of the
Bayesian analysis. Additionally, in most applications, Vt and Wt are unknown, so prior specification
of these model parameters is required to complete the Bayesian model formulation. are often used
for priors on Vt and Wt (West and Harrison, 1997).
Inference about linear combinations of the state vector (θt )t≥1 is of primary interest for creating
smooth representations of τ . Using the DLM context, inference about the state vector at time s from
the posterior distribution π(θs |y1:t ) is threefold: filtering (when s = t), forecasting (when s > t),
and smoothing (when s < t). In filtering problems data arrive sequentially in time, and the goal is
to estimate the current value of the state vector given data up to that point in time. We will use
filtering to compute the most up to date estimate of τ and quantify the existing noise. Forecasting
involves considering all the data available up to time t and making predictions about the state at
future time points s = t + k, k = 1, 2, . . . . We will use forecasting to predict the relevant minimum
and maximum values of τ for the current growing season, along with the associated prediction error.
Finally, smoothing is a retrospective problem used to study the historical evolution of the system
θ0:t using all the data available up to time t. We will use smoothing to deliver post-season estimates
of the timing of local extrema and the associated noise level.
One of our goals in this proposal is to exploit the DLM representation of the response τ so that
it incorporates the seasonal pattern, change in variability over time, and relationship to covariates
(such as weather related variables). We provide here an example of a DLM representation for
another set of data, which closely resembles the problem posed in this proposal. In the example
below, the response is an index of greenness, MTCI, collected from MERIS, on a nearly weekly
basis (8–day composites), for a period of 5 years (a total of 230 time points), over a 10×10 pixels
region in Southern India. The DLM representation proposed for the data collected in each pixel
assumes the mean is modeled as a random walk, and seasonality modeled using the Fourier basis
representation, such as:
Yt = µt + S1t + S2t + vt
µt = µt−1 + wt
Sjt =
∗
Sjt
=
vt ∼ N (0, σe2 )
(1.9a)
N (0, σl2 )
(1.9b)
wt ∼
∗
cos(tωj )Sj,t−1 + sin(tωj )Sj,t−1
+ ut
∗
−sin(tωj )Sj,t−1 + cos(tωj )Sj,t−1
+ u∗t
θ0 ∼ N (m0 , C0 )
(1.9c)
ut , u∗t ∼ N (0, σs2 )
(1.9d)
(1.9e)
ind
σe , σl , σs ∼ C + (0, 0.01)
(1.9f)
∗ , S , S ∗ ) and ω = 2πj for harmonics j = 1, 2 with a period s = 46
where θt = (µt , S1t , S1t
2t
j
2t
s
∗ =
corresponding to the 46 8–day composite values per year. Sjt = aj cos(tωj ) + bj sin(tωj ) and Sjt
−aj sin(tωj ) + bj cos(tωj ). vt , wt , ut , and u∗t are assumed independent of each other and for all
t = 1, . . . , 230. Note that we specified half Cauchy distribution priors on the standard deviations σ
to allow appropriately small values for the variance
parameters.
This model can be written in the
2
general DLM framework with Vt = σe , Ft = 1 0 1 0 1


1
0
0
0
0
0 cos(ω1 t) sin(ω1 t)

0
0




0
0
Gt = 0 −sin(ω1 t) cos(ω1 t)

0
0
0
cos(ω2 t) sin(ω2 t) 
0
0
0
−sin(ω2 t) cos(ω2 t)
13
Irrigated Agriculture
MTCI
2
1
0
−1
2003
2004
2005
2006
2007
2008
2007
2008
Time
Coastal Vegetation
1.0
MTCI
0.5
0.0
−0.5
2003
2004
2005
2006
Time
Figure 1.6: DLM results for two separate pixels for the phenology data (grey) along with the
smoothed representation obtained from the fully Bayesian analysis (red lines) and 95% credible
intervals (blue dotted lines).

 2
σl 0 0 0 0
 0 σs2 0 0 0 


2

Wt = 
 0 0 σs 0 0 
 0 0 0 σs2 0 
0 0 0 0 σs2
Under this formulation, we allow both the mean and the seasonal state components to evolve
over time and the seasonal variance parameter σs2 can be used as an indicator of the evolution of
the phenological metrics over time. For example, a very small value of σs2 (close to 0) suggests the
seasonal component does not evolve much across years, whereas a larger value of σs2 suggests the
phenological behavior may have changed across years. The fully Bayesian specification of the DLM
allows us to provide a measure of uncertainty in these estimates using draws from the posterior
(i)
distribution, π(θ0:T , ψ|y1:T ). For i = 1, . . . , M draws of the posterior distribution, compute S1:T and
then the relevant indices, such as local minimum and maximum for each growing season. The form
of the posterior is analytically intractable, so we simulate the posterior distribution of the states
and variance components, π(θ0:T , ψ|y1:T ), using MCMC, where ψ = (σe , σl , σs ). We employed a
two-step Gibbs sampling algorithm to simulate from the posterior.
1. Draw states θ0:T from π(θ0:T |ψ, y1:T ) using forward filtering backward sampling (FFBS).
2. Draw σe , σl , σs from π(ψ|θ0:T , y1:T ) using Metropolis-Hastings algorithms with InverseGamma proposal distributions.
Some of the results of a fully Bayesian analysis are summarized by the graphs provided in
Figure 1.6. We can note, for example, that the timing associated with maximum greenness varies
14
across pixels (surely related to land cover) as well as from year to year. We also note that the
variability associated with the smooth estimates of greenness are not constant throughout the time
period considered. The analysis provided information about all the variance parameters, as well as
forecasts for one-year ahead (results not shown due to space limitations).
1.6.2
Objective #3: Higher–quality Post–season Data
Another goal of the proposed work is to extend the DLM approach from modeling each of the pixels
independently to a model that incorporates spatial dependence among neighboring pixels. To our
knowledge, this is novel methodology, and we propose to approach this problem by imposing a
spatial structure on the various components in the DLM specification, such as, for example, the
mean structure or the seasonality components. Our first step will be to model the spatial field
of the DLM mean as a conditionally specified Markov Random Field, with a first or second order
neighborhood structure. In addition, we can, without much difficulty, allow for covariates to enter
the currently constant mean structure in the model specification for the phenology example.
We will also work with Collaborators Brown and Misra from NASA Jet Propulsion Laboratory
to enhance the quality of post–season data. Brown and Misra will use time as an extra dimension
to apply a combination of “super–resolution” and spectral enhancement to increase the spatial
resolution of SMAP TB . The basic theory of the super–resolution is the use of multiple pixels
or images to fill in the spectral gaps (or under–sampling) of a single image and thereby achieve
a higher possible spatial resolution. Multiple images of the same location will help reduce noise.
Since a high resolution product is preferable around the period of peak τ the dimensionality of the
algorithm (in terms of resolution and time) can be modified in an optimal way. We will use the
enhanced TB to retrieve our own version of τ using the DCA and/or MPRA.
1.6.3
Objectives #4, #5, and #6: Illustrate Value; Identify New Agronomic
Signals; Construct a Long–term Record
We will illustrate the value of our new SMAP τ product in the Corn Belt in two ways. First, we
will compare the day of year of peak τ during the growing season to USDA estimates of when corn
has reached the R3 phenological stage. that are provided in weekly Crop Progress and Condition
Reports (http://www.nass.usda.gov) for Crop Reporting Districts like those illustrated in Figure 1.5. Second, we will compare the day of year of peak τ during the growing season with the
day at which 1000 ◦ C day GDD have been reached after planting by using USDA estimates of
planting dates and meteorological data from either NWS COOP stations or other data deemed to
be superior. We will make this information available on the web and publicize it broadly with the
help of extension faculty at Iowa State University and the SMAP mission.
We will work to identify new agronomic signals in our new product for major agricultural regions
of the world using the qualitatively understanding of what τ represents in agricultural systems as
illustrated in Figure 1.1. We will initially target wheat grown in Canada and Eastern Europe and
corn and soybeans in South America. Besides the peak in τ during the growing season, there are
other interesting features. For example, there is a local maxima in the early spring and late fall
that we believe are related to tillage which roughens the soil surface and is interpreted by SMAP
as an increase in τ . There are also local minima later in the spring and near harvest. We will work
to identify yield data and records of agricultural management collected by the USDA and agencies
of foreign governments that could allow us to relate these inflection points in τ time series to other
crops and specific agricultural practices.
Finally, we will use SMAP τ along with SMOS τ to create a 10–year record of τ in the major
15
agricultural regions of the world. As long as SMOS continues to operate we will compare overlapping
records of SMOS Level 2 τ with SMAP DCA/MPRA τ . We will also construct our own custom
SMOS τ product using SMOS Level 3 TB estimated at the θ = 40◦ incidence angle of SMAP, to see
whether this custom τ compares more favorably with SMAP DCA/MPRA τ than SMOS Level 2 τ .
We will use CDF matching in order to merge the two data sets.
1.7
Plan of Work
We propose a three–year project. PI Hornbuckle will lead the project and supervise one of the
graduate students, Victoria Walker, who will defend her M.S. thesis “Why is SMOS Dry Compared
to Soil Moisture Observed by the South Fork In Situ Soil Moisture Network” in March. Ms. Walker
attended a week–long training on the use of SMOS data held in conjunction with the 2nd SMOS
Science Conference in May 2015 and is adept at manipulating SMAP data. Co–I Caragea will
supervise the other graduate student who is yet to be identified.
We expect the following work schedule.
Year 1: Jun 2016 – May 2017 The PI and Walker will deliver SMOS and SMAP τ to the Co–I
and the other graduate student, and begin comparing SMOS and SMAP τ in order to produce
the merged record. The Co–I and the other graduate student will develop an appropriate
DLM representation of τ for each pixel in the initial region of interest. Developing the model
includes proper construction of a Bayesian framework, investigation of appropriate priors,
convergence issues, and interpretation of the results and assessment.
Year 2: Jun 2017 – May 2018 The PI and Walker will begin to identify potential agronomic
signals in τ and set up the framework to create our own retrieved values of τ from TB supplied
by Collaborators Brown and Misra. The Co–I and the other graduate student will extend the
analysis to a larger geographical region, finding computationally efficient algorithms if needed.
They will also conduct the initial investigation into incorporation of spatial dependence within
the DLM model.
Year 3: Jun 2018 – May 2019 The PI and Walker will complete the merging of SMOS and
SMAP τ , document the agronomic signals that have been identified, and compare signals in
the Corn Belt with USDA estimates. The Co–I and the other graduate student will complete
the incorporation of spatial dependence, add covariates to the basic DLM developed in the
first year, and prepare a user–friendly product that can be used by a larger audience. The
PI and Co–I will prepare an algorithm theoretical basis document (ATBD) for the product.
Journal publications will be written.
1.8
Data Sharing Plan
We will share the validation data generated via this project with others on a site associated with
the Iowa Environmental Mesonet similar to the site where we currently archive SMOS data for the
U.S. Midwest (http://mesonet.agron.iastate.edu/smos/). Since netCDF format has evolved
as the de facto standard for weather and climate research, we will make our data available in this
format. Data will be made available after we have published the first refereed paper using the data
in question or after the end of the project, whichever comes earlier. This will ensure that we have
sufficient experience with the data to evaluate its validity and completeness.
16
Bibliography
Abendroth, L. J., R. W. Elmore, M. J. Boyer, and S. K. Marlay (2011), Corn growth and development, Tech. Rep. PMR 1009, Iowa State University Extension, Ames, IA.
Archontoulis, S. V., F. E. Miguez, and K. J. Moore (2014), Evaluating APSIM maize, soil water,
soil nitrogen, manure, and soil temperature modules in the Midwestern United States, Agron. J.,
106 (3), 1025–1040, doi:10.2134/agronj2013.0421.
Campbell, G. S., and J. M. Norman (1998), An Introduction to Environmental Biophysics, Springer–
Verlag, New York.
Escorihuela, M. J., A. Chanzy, J.-P. Wigneron, and Y. H. Kerr (2010), Effective soil moisture
sampling depth of L–band radiometry: A case study, Remote Sens. Environ., 114 (5), 995–1001,
doi:10.1016/j.rse.2009.12.011.
Holmes, T. R. H., M. Drusch, J.-P. Wigneron, and R. A. M. de Jeu (2008), A global simulation
of microwave emission: Error structures based on output from ECMWF’s operational integrated
forecast system, IEEE Trans. Geosci. Remote Sens., 46 (4), 846–856, doi:10.1109/TGRS.2007.
914798.
Hornbuckle, B. K., J. C. Patton, A. VanLoocke, A. E. Suyker, M. C. Roby, V. A. Walker, E. R.
Iyer, D. E. Herzmann, and E. A. Endacott (2016), SMOS optical thickness changes in response
to the growth and development of crops, crop management, and weather, Remote Sens. Environ.,
in review.
Hunt, K. P., J. J. Niemeier, L. K. da Cunha, and A. Kruger (2011), Using cellular network signal
strength to monitor vegetation characteristics, IEEE Geosci. Remote Sens. Lett., 8, 346–349,
doi:10.1109/LGRS.2010.2073677.
Jackson, T. J., and T. J. Schmugge (1991), Vegetation effects on the microwave emission of soils,
Remote Sens. Environ., 36, 203–212.
Jones, M. O., L. A. Jones, J. S. Kimball, and K. C. McDonald (2011), Satellite passive microwave
remote sensing for monitoring global land surface phenology, Remote Sens. Environ., 115, 1102–
1114, doi:10.1016/j.rse.2010.12.015.
Kerr, Y. H., et al. (2010), The SMOS mission: New tool for monitoring key elements of the global
water cycle, Proc. IEEE, 98 (5), 666–687, doi:10.1109/JPROC.2010.2043032.
Kerr, Y. H., et al. (2012), The SMOS soil moisture retrieval algorithm, IEEE Trans. Geosci. Remote
Sens., 50 (5), 1384–1403, doi:10.1109/TGRS.2012.2184548.
17
Levis, S., G. B. Bonan, E. Kluzek, P. E. Thornton, A. Jones, W. J. Sacks, and C. J. Kucharik
(2012), Interactive crop management in the Community Earth System Model (CESM1): Seasonal
influences on land–atmosphere fluxes, J. Climate, 25, 4839–4859, doi:10.1175/JCLI-D-1100446.1.
Lobell, D. B., G. L. Hammer, G. McLean, C. Messina, M. J. Roberts, and W. Schlenker (2013), The
critical role of extreme heat for maize production in the United States, Nature Climate Change,
doi:10.1038/NCLIMATE1832.
Lobell, D. B., M. J. Roberts, W. Schlenker, N. Braun, B. B. Little, R. M. Rejesus, and G. L.
Hammer (2014), Greater sensitivity to drought accompanies maize yield increase in the U.S.
Midwest, Science, 344, 516–519, doi:10.1126/science.1251423.
Patton, J., and B. Hornbuckle (2013), Initial validation of SMOS vegetation optical thickness over
Iowa, IEEE Geosci. Remote Sens. Lett., 10 (4), 647–651, doi:10.1109/LGRS.2012.2216498.
Ramsay, J. O., H. Wickham, S. Graves, and G. Hooker (2013), FDA: Functional Data Analysis, R
package version 2.3.6.
Slatyer, R. S. (1967), Plant–Water Relationships, Academic Press, New York.
West, M., and J. Harrison (1997), Bayesian Forecasting and Dynamic Linear Models, 2 ed.,
Springer–Verlag, New York.
Wigneron, J.-P., et al. (2007), L–band microwave emission of the biosphere (L–MEB) model: Description and calibration against experimental data sets over crop fields, Remote Sens. Environ.,
107, 639–655, doi:10.1016/j.rse.2006.10.014.
Zobeck, T. M., and C. A. Onstad (1987), Tillage and rainfall effects on random roughness: A
review, Soil & Tillage Research, 9, 1–20.
18
Chapter 2
Investigator Curriculum Vitae
19
Brian K. Hornbuckle
Education
Ph.D.
The University of Michigan, Electrical Engineering and Atmospheric Science
(Geoscience and Remote Sensing), 2003.
M.S.E.
The University of Michigan, Electrical Engineering (Electromagnetics), 1997.
M.A.
The University of Mississippi, Secondary Education (Science), 1996.
Sc.B.
Brown University, Electrical Engineering (Systems), 1994.
Appointments
2009 –
Associate Professor, Department of Agronomy, Department of Electrical and
Computer Engineering (courtesy), Department of Geological and Atmospheric
Sciences (courtesy), The Iowa State University of Science and Technology.
2003 – 2009
Assistant Professor, The Iowa State University of Science and Technology.
Significant Publications
Rondinelli, W. J., B. K. Hornbuckle, J. C. Patton, M. H. Cosh, V. A. Walker,
B. D. Carr, and S. D. Logsdon, Different Rates of Soil Drying After Rainfall
are Observed by the SMOS Satellite and the South Fork In Situ Soil Moisture
Network, Journal of Hydrometeorology, doi:10.1175/JHM-D-14-0137.1, 2015.
Patton, J. and B. Hornbuckle, Initial Validation of SMOS Vegetation Optical Thickness in Iowa, IEEE Geoscience and Remote Sensing Letters, doi:
10.1109/LGRS.2012.2216498, 2013.
Rowlandson, T. L., B. K. Hornbuckle, L. M. Bramer, J. C. Patton, and S. D.
Logsdon, Comparisons of Evening and Morning SMOS Passes over the Midwest
United States, IEEE Transactions on Geoscience and Remote Sensing, doi:
10.1109/TGRS.2011.2178158, 2012.
Du, J., T. J. Jackson, R. Bindlish, M. H. Cosh, L. Li, B. K. Hornbuckle, and
E. D. Kabela, Effect of dew on aircraft–based passive microwave observations
over an agricultural domain, Journal of Applied Remote Sensing, doi:10.1117/
1.JRS.6.063571, 2012.
Wigneron, J.–P., Y. Kerr, P. Waldteufel, K. Saleh, M.–J. Escorihuela, P. Richaume,
P. Ferrazoli, P. de Rosnay, R. Gurney, J.–C. Calvet, J. P. Grant, M. Guglielmetti, B. Hornbuckle, C. Mätzler, T. Pellarin, and M. Schwank, L–band Microwave Emission of the Biosphere (L–MEB) model: Description and calibration
against experimental data sets over crop fields, Remote Sensing of Environment,
doi:10.1016/j.rse.2006.10.014, 2007.
Hornbuckle, B. K., and A. W. England, Radiometric sensitivity to soil moisture at 1.4 GHz through a corn crop at maximum biomass, Water Resources
Research, doi:10.1029/2003WR002931, 2004.
1
20
Hornbuckle, B. K., A. W. England, R. D. De Roo, M. A. Fischman, and D. L.
Boprie, Vegetation canopy anisotropy at 1.4 GHz, IEEE Transactions on Geoscience and Remote Sensing, doi:10.1109/TGRS.2003.817192, 2003.
Other Publications
Bramer, L. M., B. K. Hornbuckle, and P. C. Caragea, How many measurements
of soil moisture within the footprint of a ground–based microwave radiometer are required to account for meter–scale spatial variability? Vadose Zone
Journal, doi:10.2136/vzj2012.0100, 2013.
Franz, T. E., M. Zreda, R. Rosolem, B. K. Hornbuckle, S. L. Irvin, H. Adams,
T. E. Kolb, C. Zweck, and W. J. Shuttleworth, Ecosystem scale measurements
of biomass water using cosmic–ray neutrons, Geophysical Research Letters, doi:
10.1002/grl.50791, 2013.
Cosh, M. H., E. D. Kabela, B. Hornbuckle, M. L. Gleason, T. J. Jackson, and
J. H. Prueger, Observations of dew amount using in situ and satellite measurements in an agricultural landscape, Agricultural and Forest Meteorology,
doi:10.1016/j.agrformet.2009.01.004, 2009.
Kabela, E. D., B. K. Hornbuckle, M. H. Cosh, M. C. Anderson, and M. L.
Gleason, Dew frequency, duration, amount, and distribution in corn and soybean during SMEX05, Agricultural and Forest Meteorology, doi:10.1016/j.
agrformet.2008.07.002, 2009.
Robinson, D. A., C. S. Campbell, J. W. Hopmans, B. K. Hornbuckle, S. B.
Jones, R. Knight, F. Ogden, J. Selker, and O. Wendroth, Soil moisture measurement for ecological and hydrological watershed–scale observatories: A review,
Vadose Zone Journal, doi:10.2136/vzj2007.0143, 2008.
Scientific, Technical, and Management Performance
2009 –
Member, NASA SMAP Cal/Val and Algorithm working groups.
2009 –
Associate Editor, Advances in Water Resources.
2005 –
Member, Validation and Retrieval Team, European Space Agency Soil Moisture
and Ocean Salinity (SMOS) mission.
2007 – 2012
PI, NASA Grant NNX08AE48G from the Terrestrial Hydrology Program, Quantifying the effect of intercepted precipitation and dew on terrestrial microwave
emission at L–band. Ph.D. student graduated with one accepted publication
and four manuscripts in preparation.
2006 – 2012
PI, NASA Grant NNG06GC63G from the Terrestrial Hydrology Program, A
prototype remote sensing validation site: Towards a multi–variable approach to
validating and scaling remotely–sensed observations of the water cycle. Managed a team of four Co–I’s from Iowa State and the University of Iowa to
establish the Iowa Validation Site and work towards NASA SMAP validation.
Contact Information
3007 Agronomy Hall, Ames, IA 50011-1010, bkh@iastate.edu
2
21
Petruţa C. Caragea
Education
Ph.D. Statistics
M.S. Statistics
B.S. Applied Mathematics
Appointments
Associate Chair
Visiting Faculty
Associate Professor
Assistant Professor
2003
2002
1997
2014–present
Fall 2010
2009–present
2003–2009
University of North Carolina, Chapel Hill, NC
University of North Carolina, Chapel Hill, NC
University of Bucharest, Romania
Department of Statistics, Iowa State University
Department of Applied Probability and Statistics
University of California, Santa Barbara
Department of Statistics, Iowa State University
Department of Statistics, Iowa State University
Significant Related Publications
1. Caragea, P.C., Berg, E.J., 2014. A Centered Bivariate Spatial Regression Model for Binary Data
with an Application to Presettlement Vegetation Data in the Midwestern United States. Journal of
Agricultural, Biological and Environmental Statistics, 19(4), 453–471.
2. Osthus, D., Caragea, P.C., Higdon, D., Morley, S.K., Reeves, G.D., and Weaver, B.P., 2014. Dynamic
Linear Models for Forecasting of Radiation Belt Electrons and Limitations on Physical Interpretation
of Predictive Models, Space Weather, 12, 426–446.
3. Bramer, L., Hornbuckle, B., Caragea, P.C., 2013. How Many Measurements of Soil Moisture within the
Footprint of a Ground-based Microwave Radiometer are Required to Account for Meter-scale Spatial
Variability?, Vadose Zone Journal, 12(3).
4. Kaiser, M.S., Caragea, P.C., Furukawa, K., 2012. Centered Parameterizations and Dependence Limitations in Markov Random Field Models. Journal of Statistical Planning and Inference, 142(7):
1855–1863.
5. Kaiser, M.S., Caragea, P.C., 2009. Exploring Dependence with Data on Spatial Lattices. Biometrics,
65(3): 857–865.
Scientific, Technical and Management Performance
Research Grants
1. Temporal Modeling in Integrated Predictive Analytics Methodologies for Streaming Data, Subcontract
from PNNL, PI, 2016. $21,307.
2. Hierarchical Statistical Models for the Analysis of Nn-farm Agricultural Trials of Fungicide Use in
Soybeans from Iowa Soybean Association, PI, 2014–2015. $24,683.
3. Analyses of Data Collected in On-Farm Replicated Strip Trials using Hierarchical Model and Bayesian
Statistical Analysis from Iowa Soybean Association, PI, 2011–2012. $27,806.
4. Wind Forecast Model Validation and Improvement for the Central US from U.S. Department of Energy,
CoPI , 2009–2010. $250,000.
5. Statistical Analysis for the Effects of Management Practices, Soil, and Weather on Corn Yield Differences in Normal Minus 50 On Farm Nitrogen Trials in Iowa from Iowa Soybean Association, PI,
2009–2010. $38,160.
Elected Positions & Editorial Appointments
1. Associate Editor, Journal of Agriculture, Biological and Environmental Statistics, 2011–2015.
2. American Statistical Association, Section on Statistics and Environment: Past Chair (2014), Chair
(2013), Program Chair (2010).
3. American Statistics Association, Committee on Award of Outstanding Statistical Application: Chair:
2010–2012.
Contact Information 1121 C Snedecor Hall, Ames, IA 50011 (pcaragea@iastate.edu)
1
22
Chapter 3
Current and Pending Support
PI Brian K. Hornbuckle
current “Validation of Satellite Observations of Soil Moisture to Facilitate Forecasts of Soil Water
Storage in Iowa.” PI: Hornbuckle. Program: Iowa Water Center (Rick Cruse and Melissa
Miller, millerms@iastate.edu). Period: 6/2014 to 5/2016. Budget: $115,140. Hornbuckle
commitment: 1 month per year.
pending “Calibrated Long-term Records of Soil Moisture, Vegetation Water, and Soil Surface
Roughness in the U.S. Corn Belt Using Satellite Lband Radiometry.” PI: Hornbuckle.
Program: NASA ROSES 2015 A34 NNH15ZDA001N-SCIS Satellite Calibration Interconsistency Studies (Lucia Tsaoussi, lucia.s.tsaoussi@nasa.gov). Period: three years. Budget:
$384,982. Hornbuckle commitment: 1 month per year.
pending “Analysis and Synthesis of Data Sources for Soil Moisture Estimation and Mapping.”
CoI’s: Hornbuckle and Amy Kaleita. Program: John Deere Company (Thomas Mueller,
MuellerThomasG@JohnDeere.com). Period: two years. Budget: $153,207. Hornbuckle commitment: 0.5 month per year.
Co–I Petruţa Caragea
current “Temporal Modeling in Integrated Predictive Analytics Methodologies for Streaming
Data.” PI: Caragea. Program: Subcontract from PNNL. Period: 2016. Budget: $21,307.
Caragea commitment: 1 month.
23
Chapter 4
Proposal Budget
4.1
Budget Narrative
The proposed start date is June 1, 2016.
4.1.1
Direct Labor
PI Hornbuckle is an associate professor in the Department of Agronomy at Iowa State University.
One month of salary support from NASA is requested for each budget year. He will coordinate the
project and directly supervise one of the two graduate students who will work on the project.
Co–I Caragea is an associate professor in the Department of Statistics at Iowa State University.
One month of salary support from NASA is requested for each budget year. She will directly
supervise one of the two graduate students who will work on the project.
The graduate students will be supported 12 months of the year, each year of the project, on
half–time research assistantships traditionally used in academia. This type of employment assumes
that each student will spend roughly half of their time fulfilling the obligations detailed in the
proposal and the other half fulfilling their personal obligations (for which they are not monetarily
compensated) which include academic courses and thesis preparation.
The fringe benefits are calculated at 31.5% for faculty and 13.0% for graduate students.
Salaries are incremented by 3% in the second and third years of the project.
4.1.2
Direct Costs: Equipment
No funds for equipment are requested.
4.1.3
Direct Costs: Travel
Domestic
One trip is budgeted each year to cover travel expenses for the PI to attend SMAP Cal/Val meetings.
This trip is estimated to cost $1000 based upon similar trips to Oxnard, CA, and Pasadena, CA,
for SMAP Cal/Val meetings in 2011 ($950), 2012 ($1120), and 2014 ($1000).
Trips for the PI, Co–I, and one of the graduate students to a domestic science conference (e.g.
Fall Meeting of the American Geophysical Union in San Francisco, CA) are included in the second
and third years of the budget. The purpose of these trips is to present new project results to the
scientific community as they become available and to collaborate with other scientists. These trips
are estimated to cost $2100 for the PI and Co–I, and $1900 for a graduate student, based upon
24
previous trips. In the second year there is one trip for the PI and one for the Co–I. In the third
year there is one trip for the Co–I and one trip for a graduate student.
Foreign
One trip for the PI and one trip for one of the graduate students is budgeted in the third year
of the project to cover travel expenses to attend the IEEE International Geoscience and Remote
Sensing Symposium (IGARSS). The purpose of these trips is to present new project results to the
scientific community as they become available and to collaborate with other scientists. IGARSS
will be held in Spain in 2018. The trip for the PI is estimated to cost $3200 (the PI paid $3150
in 2009, $3060 in 2010, $2800 in 2011, and $3200 in 2012) and the trip for the graduate student
$2800.
4.1.4
Other Direct Costs
Materials and Supplies
None.
Publication Costs
Funds are requested in the second and third year of the project to help to defray publication costs
of scientific articles in various peer–reviewed journals as a result of this research.
Equipment or Facility Fees
None.
Other
Funds are requested for tuition support for the two half–time research assistantships at Iowa State
University for each budget year.
4.1.5
Indirect Costs
Iowa State University’s DHHS negotiated on–campus indirect cost rate is 50% of total direct costs
excluding equipment, tuition allowance, patient care costs, and subcontract amounts exceeding
$25,000.
4.2
Proposal Personnel and Work Effort
Table 4.1 summarizes the personnel required to complete the project described in this proposal,
and the work efforts that will be dedicated to the project if this proposal is successful. The PI and
Co–I request that NASA provide funds equivalent to one month of their time. The PI and Co–I also
expect to devote approximately one–third of their research time during the academic year to this
project. Collaborators Misra and Brown of NASA Jet Propulsion Laboratory will work to improve
the spatial resolution of SMAP data on their own NASA project (if funded), and will provide the
resulting data to this project.
25
Table 4.1: Names and planned work commitments of project personnel.
Personnel
Work Commitment
Brian Hornbuckle (PI)
1 month per year (NASA funds) for 36 months plus additional
time as a result of his research appointment at Iowa State
Petruţa Caragea (Co–I)
1 month per year (NASA funds) for 36 months plus additional
time as a result of her research appointment at Iowa State
(Graduate Student)
20 hours/week for 36 months (6 person–months each year)
(Graduate Student)
20 hours/week for 36 months (6 person–months each year)
Sid Misra (Collaborator)
time necessary to provide refined SMAP data
Shannon Brown (Collaborator)
time necessary to provide refined SMAP data
4.3
Facilities and Equipment
Office space and general computing equipment are provided by the Departments of Agronomy and
Statistics at Iowa State University.
26