Phenology Week Talk October 3, 2014 Cynthia Wallace Research Geographer, U.S. Geological Survey, Tucson AZ Buffelgrass Phenology Project: When and Where.

Phenology Week TalkOctober 3, 2014

Cynthia WallaceResearch Geographer, U.S. Geological Survey, Tucson AZ

Buffelgrass Phenology Project:When and Where Buffelgrass is Green

Phenology Days Celebrationwww.usanpn.org/nn/phenoweek2014

Thanks to all of our Partners!

The Sonoran Desert Ecosystem

Wide spacing between individual plants means that fires that start do not spread

Natural fires are infrequent and of low intensity

Credit: Caroline Patrick-Birdwell

http://www.tomuphoto.com/gallery/sonoran-desert/Palo_verde_flowering_2?full=1

Buffelgrass (Pennisetum ciliare) invasion in the foothills of the Santa Catalina Mts.

north of Tucson

Credit: Caroline Patrick-Birdwell

Get it while it’s Green: Using MODIS satellite data to capture dynamics of buffelgrass (Pennisetum ciliare) phenology for eradication and management

USGS science and funds directed toward NPS management issue

Working with Saguaro National Park

Coupling field-based observation of buffelgrass phenology with satellite “land surface phenology” and climate

Model when and where buffelgrass is green

Alert managers in a timely manner to treat optimally with herbicides

Data MODIS Satellite Data

Buffelgrass Phenology Observations

SNP Mapping of Buffelgrass

Climate Data

Preliminary Results

Next Steps




Climate Data

Preliminary Results

Next Steps

Passive Sensors A passive sensor records reflected or emitted energy. Usually, the energy source is the sun.

Examples include aerial and satellite images.

Note: Photographs and image data are different. Photos are recorded on sensitized film whereas images are recorded as electronic data. From: Shane Brandt, Geospatial Extension Agent, U of New Hampshire

Digital Format & Brightness Values

Digital format: the subdivision of an image into small equal-sized and shaped areas, called picture elements or pixels, and representing the

brightness of each area with a numeric value or digital number.

Canada Center for Remote Sensing

Pixel: Sensor:

1 m IKONOS

1 m Quickbird

30 m TM, ETM+

250 m MODIS

1 km AVHRR

How a Satellite Image is Made

Energy

Data

Display

Per

cent

Ref

lect

ance

Near Infrared (0.83 um)

DA

TA

Green Gun

Green (0.56 um)

SWIRNIRB G R

Red (0.66 um)

Blue Gun Red Gun

Composing a Satellite ImageE

NE

RG

Y

DIS

PLA

Y

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 um

50

40

30

20

10

1 2 3 4 5 7TM Bands

Basalt

Vegetation

Soil

Dar

k --

----

----

----

-- B

right

$T

Per

cent R

efle

ctan

ce


DA

TA

Green Gun

Green (0.56 um)

SWIRNIRB G R

Red (0.66 um)

Blue Gun Red Gun


NE

RG

Y

DIS

PLA

Y

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 um

50

40

30

20

10

1 2 3 4 5 7TM Bands

Basalt

Vegetation

Soil

Dark

---

----

----

----

- B

right

$T

Energy

Blue band

Green

band

Red band

NIR

band

A band of an image is the measurements taken in a specific region of the electromagnetic spectrum.

Per

cen

t R

efle

ctan

ce


DA

TA

Green Gun

Green (0.56 um)

SWIRNIRB G R

Red (0.66 um)

Blue Gun Red Gun

Composing a Satellite Image

EN

ER

GY

DIS

PLA

Y

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 um

50

40

30

20

10

1 2 3 4 5 7TM Bands

Basalt

Vegetation

Soil

Da

rk -

----

----

----

---

Bri

ght

$T

Data (from 3 bands of energy)

In this case, we are creating a “false color” composite, where by convention, green band energy is assigned the color blue, red band

energy is assigned the color green, and near infrared energy (which is directly undetectable by the human eye) is assigned the color red,

Per

cent

Ref

lect

ance


DA

TA

Green Gun

Green (0.56 um)

SWIRNIRB G R

Red (0.66 um)

Blue Gun Red Gun

Composing a Satellite Image

EN

ER

GY

DIS

PLA

Y

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 um

50

40

30

20

10

1 2 3 4 5 7TM Bands

Basalt

Vegetation

Soil

Da

rk -

----

----

----

---

Bri

ght

$T

Display – a “False Color Composite”

Near-IR (band 4) Red (band 3) Green (band 2)Blue (band 1)

SATELLITE

COMPUTER

From: Shane Brandt, U of New Hampshire

Per

cent R

efle

ctan

ce


DA

TA

Green Gun

Green (0.56 um)

SWIRNIRB G R

Red (0.66 um)

Blue Gun Red Gun


NE

RG

Y

DIS

PLA

Y

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 um

50

40

30

20

10

1 2 3 4 5 7TM Bands

Basalt

Vegetation

Soil

Dark

---

----

----

----

- B

right

$T

Spectral Signature

Blue band

Green

band

Red band

NIR

band

The pattern of reflectance characteristic of a surface. Vegetation has a distinctive spectral signature.

Plant Phenology and Spectral Response – One Species

• Chlorophyll in green leaves absorbs blue and red, reflects green and NIR

• Senescent leaves reflect more red, less NIR

• Band ratio can distinguish actively photosynthesizing vegetation from senescent

• Observing seasonal patterns can identify plant species, guide management decisions

(From: Jensen, John R. Remote Sensing of the Environment: An Earth Resource Perspective.)

Vegetation Indices

V.I.s are Remote sensing surrogates for

estimating certain biophysical parameters of

plants (e.g., percent cover, biomass, LAI)

From: Paul Pinter,USDA, ARS, US Water Conservation Laboratory

For example:Normalized Difference Vegetation Index

NDVI = (NIR - Red)/(NIR + Red)




Climate Data

Preliminary Results

Next Steps

Ventana Trail Site Ventana Waterfall Site Ventana Mist Site
















Areas in red outline patches of dense buffelgrass from 2008 aerial mapping of the western Santa Catalina foothills area of Tucson AZ.




Climate Data

Preliminary Results

Next Steps

Saguaro National Park

The spatial pattern of rainfall is highly variable in the Sonoran Desert

Photo Credit: Zack Guido, CLIMAS, The University of Arizona




Climate Data

Preliminary Results

Next Steps

Get it while it’s Green: Preliminary Results

Observed buffelgrass greenness, MODIS-EVI and Precipitation (PPT) data for 16-day periods

10-1

310

-17

10-2

111

-02

11-0

611

-10

11-1

411

-18

11-2

212

-03

12-0

712

-11

12-1

512

-19

12-2

313

-04

13-0

813

-12

13-1

613

-20

1000

1200

1400

1600

1800

2000

2200

2400

2600

2800

3000

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

MODIS-EVI

Pct_green

10-1

310

-17

10-2

111

-02

11-0

611

-10

11-1

411

-18

11-2

212

-03

12-0

712

-11

12-1

512

-19

12-2

313

-04

13-0

813

-12

13-1

613

-20

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

0

0.5

1

1.5

2

2.5

3

3.5

Pct_green

PPT Total (in)

2011 2012 2013

2011 2012 2013

Get it while it’s Green: Statistical correlation analysis

Fourier Harmonics of Annual Phenology

Fourier Harmonic Analysis of a Waveform

Forest Shrubland Grassland Wetland Forest Shrubland Grassland Wetland

1998 1999

Temporal NDVI profiles for selected SE Arizona landscapes

AZ phenometrics for 1998

Get it while it’s Green: Harmonics analysis of each variable by year

2011

• MODIS and Greenness patterns align• PPT pattern precedes Greenness

Get it while it’s Green: Harmonics analysis comparing variables

• Timing of PPT always earlier, Timing of Greenness and MODIS typically coincide

Preliminary Results

Buffelgrass greenness observed on the ground is strongly correlated to contemporaneous MODIS-EVI greenness.

Buffelgrass greenness observed on the ground is highly correlated to precipitation of the prior time-period and the prior two time periods (prior 16 day total and prior 32 day total).

Annual harmonics resonance between greenness-MODIS-PPT show consistent patterns, with PPT peaks preceding others and with greenness and MODIS in synchronicity.

Saguaro National Park: so far, we have found lower MODIS-EVI values for some vegetation types with buffelgrass invasion vs. areas without buffelgrass.




Climate Data

Preliminary Results

Next Steps

1 2 3 4 5 6 7 8 9 10 11 120

1

2

3

4

5

2002 (7.47") Ave(11.28")

2002

1 2 3 4 5 6 7 8 9 10 11 120

1

2

3

4

5

2005 (11.55") Ave(11.28")

Precipitation (UA Station)

2005

MODIS NDVI Profiles

1. Join Nature’s Notebook

2. Find a site to monitor buffelgrass – Choose a location that you will visit once every week or two. I can help you select a site or you can have your own site added to the network.

3. Sign up as an observer – Become an official participant with Nature’s Notebook and set your username and password. All you need is an email address and Internet access. When you are registering, identify yourself as part of this effort by selecting “Buffelgrass Monitoring Network, Tucson” from the Partner Organization drop-down menu.

How to participate…..

Thank you!

Phenology Week Talk October 3, 2014 Cynthia Wallace Research Geographer, U.S. Geological Survey, Tucson AZ Buffelgrass Phenology Project: When and Where.

Documents

image data

digital image

electronic data

digital number

buffelgrass mitigationjune

satellite images

phenology week talkoctober

madeenergy data display