Remote Sensing for Agricultural Applications - PowerPoint PPT Presentation

Loading...

PPT – Remote Sensing for Agricultural Applications PowerPoint presentation | free to download - id: a62aa-YjhhN



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Remote Sensing for Agricultural Applications

Description:

Remote Sensing for Agricultural Applications – PowerPoint PPT presentation

Number of Views:8414
Avg rating:3.0/5.0
Slides: 136
Provided by: igettD
Learn more at: http://igett.delmar.edu
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Remote Sensing for Agricultural Applications


1
Remote Sensing for Agricultural Applications
  • Tom Loveland and Eric Wood
  • U.S. Geological Survey / EROS

2
Examples of my USGS Participation in Agricultural
Related Projects
  • Africa
  • Sustainable Tree Crops
  • Famine early warning
  • Natl. land certification
  • Env. monitoring systems
  • Modeling land use /cover
  • Other International
  • Bananas Dominica
  • Coffee LA/Carib
  • Domestic
  • Tribal LULC N. Great Plains
  • SD rangeland monitoring
  • Chemical exposure cancer study with NCI
  • EEAP USGS Central Region

3
CROPLAND
4
CROPLAND
  • Cultivated systems, including croplands,
    shifting cultivation, confined livestock
  • production and freshwater aquaculture cover
    approximately 24 of total land
  • area
  • In the last two decades, the major areas of
    cropland expansion were located in Southeast
    Asia, parts of South Asia, the Great Lakes region
    of eastern Africa, the Amazon Basin, and the U.S.
    Great Plains
  • The major decreases of cropland occurred in the
    southeastern United States, eastern China, and
    parts of Brazil and Argentina
  • According to FAO estimates, 1 500 million
    hectares (3 706 million acres) of the worlds
    land is currently being used for growing crops
  • Of all human activities, agriculture consumes
    the greatest amount of water, accounting for 70
    of all water withdrawals worldwide

Millennium Ecosystem Assessment Ecosystems and
Human Well-being 2005 UNEP 2005
5
Greenhouse agriculture in Almeria, Spain
  • 1987 The landscape reflects rather typical
    rural agricultural land use
  • 2000 The area has been converted to intensive
    greenhouse agriculture
  • 2004 Greenhouse-dominated land appears as
    whitish gray patches

6
Converting forests into farms in Santa Cruz,
Bolivia
  • 1975 Forested landscape
  • 2003 Large corporate agricultural fields
    transform the landscape

7
Novovolynsk Contrast in land use practices,
Ukraine
1988-2000 Apparent change in the approach to
land use in Ukraine, following the patterns in
Poland
8
Difference in change pattern across Tensas River
Basin, United States
  • 1972 Clearing of forest during the 1960s and
    1970s increased flooding and erosion
  • 2001 Land cover change on the Louisiana side of
    the image is striking

2 Oct 1972
2 Oct 1972
9
Greening of Al Isawiyah desertSaudi Arabia
  • 1991 Irrigation in the desert begins
  • 2000 Irrigation transforms the desert
  • 2004 Irrigation intensity increases

10
Toshka Farming the desertEgypt
These images show the desert area transformed
into vegetable plots
  • 1984-87 Image of the area before the desert
    reclamation project began in mid-1990s

n
  • 2000 Four new lakes are visible in this image,
    the faint blue-green areas around the lakes are
    newly created agricultural lands

11
Why Use Remote Sensing for Agricultural
Applications?
  • Multi-spectral capabilities are advantageous for
    assessing crop types and conditions
  • Repetitive coverage captures seasonal changes in
    crop development
  • Historical coverage permits monitoring of changes
    in agricultural practices

12
Applications of Remote Sensing in Agriculture
  • Mapping the types and extent of agricultural land
  • Assessing and monitoring the condition of
    agricultural lands
  • Monitoring agricultural land change
  • Precision agriculture
  • Environmental assessments

13
Mapping Crop Extent and Types
14
(No Transcript)
15
Vegetation Spectral Signatures
16
NDVI (Normalized Difference Vegetation Index)
aka Greenness
Seasonal Metrics Start of Season End of
Season Length of Season Growing season
greenness Greenness to-date
17
Normal color
Color IR
18
Normalized Difference Vegetation Index
NDVI (Near IR Red)/(Near IR Red) Range
-1 to 1 NDVI (grass (50-5)/(505) 0.82
19
Normalized Difference Vegetation Index
NDVI (Near IR Red)/(Near IR Red) NDVI
(turf) (6-4)/(64) 0.20
Normalized Difference Vegetation Index NDVI
nearIR-vis / nearIRvis (range -1 to
1) NDVI(turf) 3-5/35 -.25
20
NDVI is seasonally dependent
Spring
Summer
21
NDVI is seasonally dependent
NDVI - March 2002
NDVI - June 2002
22
(No Transcript)
23
NDVI is responsive to vegetation condition
Drought
Non-Drought
24
MODIS 250 m Study Area
NDVI 16-day Comp.
25
(No Transcript)
26
(No Transcript)
27
National Land Cover Classification
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
AVHRR NDVI changes over time can be used to map
the unique vegetation development in cropped
areas.
32
(No Transcript)
33
Global Irrigated Area Map (GIAM) Aggregated 28
class
Note there is also a 323 disaggregated class map
of the World. See
http//www.iwmigiam.org
.we have achieved this.
34
Global Map of Rainfed Cropland Area (GMRCA)
Disaggregated 66 class
.Recently released.
35
Prototype CONUS Irrigated Lands Map
36
Irrigated Agriculture Methodology
MODIS Annual Peak NDVI
County irrigated area statistics
Land cover mask
37
Assessing Agricultural Lands Condition
38
Vegetation Canopy Dynamics Measures from 18
Years of Satellite Greenness Data
Seasonal Metrics Start of Season End of
Season Length of Season Growing season
greenness Greenness to-date
39
TIN
Time Integrated NDVI
40
National Agricultural Statistics Service (NASS)
AVHRR Vegetation Condition Product
41
National Agricultural Statistics Service (NASS)
AVHRR Vegetation Condition Percent Change Product
42
Principal Drought Monitor Inputs
CPC Daily Soil Model
USGS Streamflow
Palmer Drought Index

30-day Precip.
USDA Soil Ratings
Satellite Veg Health
43
Remote Sensing SourcesPotential Value for
Drought Decision Support
  • Radar rainfall estimates
  • Passive microwave surface soil moisture
  • Optical (visible, NIR, SWIR)
  • Vegetation Indices (NDVI, VCI, NDWI)
  • Thermal
  • Surface temperature
  • Cold cloud tops (rainfall estimates)

44
(No Transcript)
45
VegDRI Models Current Drought Status
Model Input Layers
Start of Season Anomaly
Seasonal Greenness Anomaly
VegDRI (Vegetation Drought Response Index)
Standardized Precipitation Index
Available Water Capacity
Regression Tree Modeling
Ecoregions
Irrigated Ag.
Land Cover
46
Integrating Satellite and Climate Data for
Improving Drought Monitoring
Climate-based drought indicators provide levels
of drought severity.
Satellite-derived vegetation condition provides a
relative measure of vegetation condition.
Regression Tree Modeling
Climate-based drought data provide some
explanation for the satellite vegetation
anomalies.
Satellite anomalies may be caused by drought,
flooding, late greennup, land cover conversion,
etc.
Vegetation Drought Response Index (VegDRI) Maps
47
Internet portal for data access and distribution
48
VegDRI map for July 25, 2002
Imagine how the USDM would have looked if one of
the inputs was VegDRI
49
Current Year VegDRI Products
July 16, 2007
50
Past Protype VegDRI Products
51
2007 VegDRI Map Access
http//gisdata.usgs.gov/website/Drought_Monitoring
/viewer.php
QuickViews
Dynamic Map Viewer
http//www.drought.unl.edu/vegdri/VegDRI_Main.htm
52
A County Story
Hand County, South Dakota
53
Hand County, SD
  • Total area 3,705 km2 (gt900,000 acres)
  • Top Five Commodities (from 1997 Ag census)
  • Cattle and calves
  • Wheat
  • Corn (for grain)
  • All other grains
  • Hogs and pigs

54
Hand County, SD
July 25, 2002
Land Cover
168,000 acres of grassland impacted by
drought 213,000 acres of row crops impacted by
drought
55
Hand County, SD
Land Cover (from National Land Cover
Database) Pink Row Crops Burgundy Small
Grains Tan Grassland Yellow Pasture/Hay
56
Hand County, SD
May 30, 2002
July 25, 2002
September 5, 2002
47 of county is normal 53 of county is in
drought
57
Monitoring Agricultural Land Change
58
1974
1992
2000
Meade County Land Change 1974 to 2000 Almost
all of this 6 by 6 square mile area in north
central Meade County was grassland in 1974. By
1992, approximately 15 percent of the area was
converted to cropland (see strip cropping
patterns). By 2000, most of the cropland
reverted to grassland. The regional color
patterns show differences in grass density
related to differences in range site
characteristics and management practices.
59
Lake Thompson, SD- 1973, 1984, and 2000 (all
summer scenes)
July 5, 1973
August 13, 1984
June 30, 2000
July 5, 1973 August 13,
1984 June 30, 2000
60
Irrigation Expansion in the Western High Plains
1973
2000
61
New Agricultural Technologies, Management Styles,
and Policies
62
Irrigation Expansion
63
Conservation Reserve Program
64
(No Transcript)
65
(No Transcript)
66
Southeastern Plains Ecoregion
This ecoregion lost over 1 million hectares of
cropland, with 84 of the lost cropland converted
to forest cover.
67
Blue Ridge (2.0)
Southern Coastal Plain (24.9)
The 20 Forested Eastern Ecoregions -- overall
1973-2000 change 12.5
68
(No Transcript)
69
Primary Eastern U.S. Land Changes 1973-2000
70
Agricultural land cover declined almost
everywhere, with a net loss of 2.4 percent
(24,891 km2) between 1973 and 2000
71
Eastern United States Most Frequent Land Cover
Conversions
72
Southern Florida Coastal Plain
73
Between 1973 and 1980, 1.3 of the Southern
Florida Coastal Plain wetlands were drained for
cropland due to the Cuba sugar embargo.
74
Southern Coastal Plain
Net loss of agricultural cover 396,981 ha
75
Marshall, C.H. Jr., R.A. Pielke Sr., and L.T.
Steyaert, 2003. Crop freezes and land-use change
in Florida. Nature, 426, 29-30.
Land Cover Change Results in Colder Temperatures
and Long Freeze Periods in Key Florida
Agricultural Regions
1900-era Land Cover
Model difference in minimum temperature
Model difference in duration of freeze
temperatures
1993 Land Cover
Areas where wetlands were converted to cropland
had colder minimum temperatures and longer
freezing periods. Wetlands once held heat from
the day, often keeping area temperatures above
freezing throughout the night.
76
Western High Plains
77
(No Transcript)
78
(No Transcript)
79
1950
1978
1997
Western Expansion of Corn
Irrigated Area Western High Plains Ecoregion
Corn region
Cotton region
80
Drivers of change
  • Population growth
  • CV cities growing in the service and information
    sectors
  • Commuters from SF Bay Area and Los Angeles
  • (Relatively) affordable housing
  • Climate variability
  • Loss (temporary?) of marginal farmlands in arid
    San Joaquin Valley
  • Irrigation advances
  • Grapes, citrus, and nut crops at ecoregion
    periphery
  • Higher-risk crops on low productivity soils due
    in part to advances in drip irrigation practices
  • Agricultural policy
  • CRP not a factor
  • California Farmland Conservancy Program
  • Waterfowl habitat programs to preserve and
    enhance migratory bird habitat

81
Environmental Impacts of Agriculture
82
(No Transcript)
83
Copeland et al., 1995
84
Utilization of the Landsat Archive for
Agricultural Chemical Exposure AssessmentSusa
n Maxwell and Eric Wood
85
Primary Collaborators
  • Dr. Mary Ward NIH/National Cancer Institute
  • Epidemiologist
  • Agricultural chemicals cancer (NHL, Leukemia,
    etc.)
  • Dr. Jay Nuckols, et al. Colorado State
    University
  • Environmental Engineer, Professor, Director EHASL
  • Application of RS/GIS technology in
    environmental/human health studies

86
Overview
  • Rural populations can be exposed to chemicals
    through the proximity of their residences to
    treated cropland.
  • Rural populations generally excluded from
    epidemiological studies because exposure is
    difficult to assess.

87
Overview
  • Traditional methods for collecting chemical
    exposure data
  • Questionnaires
  • Environmental samples (e.g. dust, air, water)
  • Biological samples (e.g., blood)
  • Problems with methods
  • Historical data difficult or impossible
  • People may not know
  • Sampling is costly
  • Subjective bias with questionnaires

Problem need an objective measure of chemical
exposure
88
Overview
Crop Chemical Corn Atrazine Vineyard Methyl
Bromide
ExposureMetric Concept
Land Cover Information
Chemical Use Information
Corn (Atrazine)
Spatial Linkage of data sets to determine where
chemicals applied
Residence
Vineyard Methyl Bromide
89
Overview
Land Cover information requirements for
agricultural chemical exposure assessment
90
Overview
Primary Objective Develop a methodology to
utilize the Landsat archive to derive historical
crop maps for Iowa to support the NIH/NCI
Agricultural Health Study
91
  • NLCD - more information in imagery to
    discriminate crop types.

Original Landsat Image
NLCD Map
92
1992 NLCD compared to 2000 Crop Map.
Object-oriented knowledge based classification
(eCognition)
93
Technical Approach
  • Develop an object-based rule-based image
    analysis approach

94
Technical Approach
  • Develop an object-based rule-based image
    analysis approach

Need calibrated/normalized images to support
automated classification
95
INTERNATIONAL EXAMPLES
96
(No Transcript)
97
  • http//www.pecad.fas.usda.gov/glam.cfm
  • http//www.pecad.fas.usda.gov/cropexplorer/

98
(No Transcript)
99
Crop Water Requirement Modeling in a GIS
Environment
  • USGS FEWS NET
  • EROS Data Center
  • Sioux Falls, SD 57198
  • USA

http//edcintl.cr.usgs.gov/adds/
U.S. Department of the Interior U.S. Geological
Survey
100
Objective
  • To monitor, seasonally, onset of rainfall and
    crop performance using a crop water balance model
  • To estimate production before the end of season

101
(No Transcript)
102
Water Requirement Satisfaction Index (WRSI)
WRSI f (ppt, pet, WHC, Crop Type, SOS,
EOS, LGP)
data from NOAA, generated at EDC
RFE (NOAA)
FAO soils map of the world
Kc (FAO)
103
Methods
  • NDVI data were masked to exclude non-agricultural
    areas using the SADC Regional Land Cover Database

Cultivated Area NDVI
SADC Land Cover
MODIS NDVI
104
Results- Relative NDVI Total Maize Production
per District
105
Africa Data Dissemination Service (ADDS)
http//earlywarning.usgs.gov/adds
Other FEWS NET links http//earlywarning.usgs
.gov/afghan http//earlywarning.usgs.gov/centralam
erica http//earlywarning.usgs.gov/haiti
106
First Draft Land Use/Land Cover Map, West Africa
2000
107
Tools for Assessing Land Resources (Senegal)
Current and seasonal vegetation production
(MODIS NDVI)
Images of land surface (Landsat)
Land Productivity (MODIS)
Land Use
108
Percent Woody Cover1943 40
70 1994 (woodlands) 10 - 20 1994 (cultivated)
1 2
October 1994
Niombatu, 25 October 1943
Saloum Agricultural Region
109
(No Transcript)
110
Video Image Mosaic Terroir de Soukouto
111
Land Use Map from Video Mosaic
112
Tree Density Changes in the Pastoral Region,
Northern Senegal
  • 6.3 trees shrubs/ha
  • 3.8 tons woody biomass/ha
  • 1.8 tons C in woody biomass/ha
  • 2 woody species/ha
  • (based on measurements at comparable
  • sites in the pastoral zone)
  • 8.6 trees shrubs/ha
  • 5.3 tons woody biomass/ha
  • 2.5 tons C in woody biomass/ha
  • 4 woody species/ha
  • (based on measurements at comparable
  • sites in the pastoral zone)

Corona 1968
Quickbird 2004
113
Farmlands abandoned for savannasPeanut Basin,
Senegal
Shows growing patchwork of savannas (greenish
patches)
  • 1979 Image shows farmland before being abandoned
  • 1999 Hundreds of villages are scattered
    throughout the region to enjoy fallow and grazing
    lands

114
Impact of drought and over-grazingRevane,
Senegal
  • 1965 Ancient valleys cutting through gravelly
    plateaus, with extensive bushland vegetation
  • 1999 The badland phenomenon spread extensively
    along the shallow valley slopes

115
(No Transcript)
116
Mapping Agricultural Expansion
117
(No Transcript)
118
Predictive Models of LULC Change
119
Rapid Land Cover Mapper (RLCM)Gray Tappan and
Matt CushingSAIC, USGS Center for EROS
U.S. Department of the Interior U.S. Geological
Survey
120
(No Transcript)
121
(No Transcript)
122
(No Transcript)
123
(No Transcript)
124
(No Transcript)
125
(No Transcript)
126
(No Transcript)
127
Time-Series Mapping with the RLCM
ETM, 2000
TM, 1984
MSS, 1972
128
Ethiopia Land Tenure Administration Program
Parcel demarcation with farmers
Field plots of IKONOS satellite image
129
(No Transcript)
130
(No Transcript)
131
(No Transcript)
132
Web Sites for Accessing Ag Data
  • Crop Explorer and MODIS images from Rapid
    Response
  • http//www.pecad.fas.usda.gov/cropexplorer/
  • MODIS/NDVI 250-meter Time Series Database from
    UMD
  • http//pekko.geog.umd.edu/usda/demo1/
  • Crop Explorer on ESRIs Geography Network
    (USDA-FAS)
  • http//www.geographynetwork.com/

133
Drought National Drought Mitigation Center
http//www.drought.unl.edu/ Current US Drought
Monitor http//www.drought.unl.edu/dm/monitor.htm
l USDA Agricultural Information Weekly
Weather and Crop Bulletin http//www.usda.gov/oce/
weather/pubs/Weekly National Agricultural
Statistics Service http//www.nass.usda.gov/ 200
2 Census of Agriculture http//www.nass.usda.gov
/Census_of_Agriculture/
134
Africa Data Dissemination Service (ADDS)
http//earlywarning.usgs.gov/adds
Other FEWS NET links http//earlywarning.usgs
.gov/afghan http//earlywarning.usgs.gov/centralam
erica http//earlywarning.usgs.gov/haiti
135
Questions ..
About PowerShow.com