1. Tools for Hazard Monitoring, Assessment, and Response
Landsat 8
Slide 1
So happy to be here today in this beautiful city. I especially want to thank the Minister of Agriculture, Senor Maritinez and Senor de la Vega, the Director General of SIAP, for extending the kind invitation to present here today. Also I look forward to fostering a closer working relationship to colleagues at SIAP as we exchange ideas this week on drought monitoring in Mexico.
First I show a few photos related to where I come from. I work for the US Geological Survey at a facility in the central plains of the US, Sioux Falls SD surrounded by corn and soybean fields. My center the Earth Resources Observation and Science Center and we are many things, but mainly we are concerned with advancing knowledge about the Earth’s land surface, how it is changing (or has changed) and developing tools for monitoring, assessment, and response. For these tools, our primary focus is on remote sensing from satellites,
We are especially glad this spring about the successful launch of the Landsat 8 mission (launched February 11). Data from that sensor should start to be available with basically global coverage this May.
Tools for Hazard Monitoring, Assessment, and Response
Jesslyn F. Brown, U.S. Geological Survey, Earth Resources Observation and Science, South Dakota, USA

2. In the face of hazards, what can people and organizations do to help our situation?
Better access to information
Responsive to a changing situation
Have the proper content, latency, spatial detail
Plan
Take actions to reduce vulnerability

3. Specific actions to address hazard management within integrated assessment and planning
An assessment of the presence and effect of natural events on the goods and services provided by natural resources
Estimates of the potential impact of hazard events on development activities
Include measures to reduce vulnerability in the proposed development activities
Drought is only one type of hazard, but it is one that dramatically effects agriculture. Generally more costly than other hazards especially when we consider the impacts and costs at multiple scales.
Organization of American States

4. Multiple tools are needed to access information about hazards
Drought is a kind of climate hazard, and it is one that dramatically effects agriculture.
Generally more costly than other hazards especially when we consider the impacts and costs at multiple scales.
Multiple tools are needed to access information about hazards
Where, When, How severe?
Remote sensing is one tool

6. Remote Sensing Advantages
Cost-effective and rapid of acquiring up-to-date information over a large geographical area
Practical way to obtain data for inaccessible places
High repetition rate and continuous coverage (time and space)
Standard methods and techniques
Remote Sensing Limitations
Indirect measure of the phenomena
Atmosphere noise needs to be corrected
Geometric and sensor calibration issues
Effect Accuracy

7. Vegetation Dynamics and VegDRI
Satellite
Systems
User/Decision
Support System
Data Services
Georegistration
Compositing
Surface Reflectance
Stacking
Smoothing
Anomaly Detection
Metrics Calculation
(SOS, SG, PASG)
Existing
AVHRR
Vegetation Dynamics
System
Data Translation
Data/products to partners >> Regular and Quick
EMODIS System
Vegetation Dynamics and VegDRI
Models
MODIS
2009>

8. Last year marked the most severe and extensive drought in at least 25 years, according to the U.S. Department of Agriculture. It was also the hottest year on record for the United States. Nearly 80 percent of farmland experienced drought in 2012, with more than 2,000 counties designated disaster areas. By September 2012, 50 percent of the crops being harvested were in poor or very poor condition.
Last year’s damaged harvest is expected to raise food prices by as much as 4 percent in 2013, particularly products like beef, which suffered from a lack of available cattle feed and viable foraging options. Overall, the 2012 drought cost an estimated $150 billion in damage, as well as an estimated 0.5 to 1 percent drop in the U.S. gross domestic product.

9. Drought Impacts
Annual direct losses to the U.S. due to drought are, on average, $6-8 billion (FEMA) $$150 billion
Drought severity can be significantly under- or over-estimated due to inadequate drought observations. This affects Planning, Prediction, Mitigation, and Response.
Impacts are evident at multiple scales (local, regional, national, global) and in multiple sectors

10. Drought Impacts
Warrick and Bowden, 1981

12. drought drought Dust Bowl drought drought drought
About 80 percent of agricultural land experiened drought in 2012, which made the 2012 drought more extensive than any since the 1950s.
The 2012 drought rapidly increased in severity from June to July and persisted into August. As of September 12, over 2,000 U.S. counties had been designated as disaster areas by USDA in 2012, mainly due to drought.
Severe or greater drought in 2012 impacted 67 percent of cattle production, and about percent of corn and soybean production.
Yields are now forecast at bushels per acre, the lowest since 1995.
Dust Bowl
drought
drought
drought

14. U.S. requirements for operational drought monitoring—geospatial products
Product coverage
National synoptic
Product schedule
Weekly, circa Monday a.m.
Spatial scale
1-4 km2, sub-county details
Product latency
<24-48 hours
Length of record (providing climatology)
~30 years for climate data, information framed related to climatology or history
These terrestrial monitoring users need data served fairly quickly but also need a consistent historical basis for determining change on the landscape.

15. Multi-year time-series observations support National monitoring of land change
Normalized Difference Vegetation Index (NDVI)
Start of Season
End of Season
Length of Season
Seasonal “greenness” –cumulative productivity
As a basis for national terrestrial monitoring, there is great value of the AVHRR satellite database that has been consistently processed and is provided by the USGS EROS Data Center as a standard product for over 20 years. We also calculate a number of other measures from the NDVI (Normalized Difference Vegetation Index—aka “Greenness”).

16. Integrating Satellite and Climate Data
Seasonal greenness condition from satellite highlights areas with anomalous vegetation condition
Deficits (compared to average condition) might be caused by drought, flooding, late greennup, land cover conversion, etc.
Climate information is needed to provide evidence that enables interpretation of the satellite data anomalies

18. Vegetation Drought Response Index Methodology
Brown, J.F., et al. (2008). The Vegetation Drought Response Index (VegDRI): A new integrated approach for monitoring drought stress in vegetation. GIScience and Remote Sensing, 45,
1. Historical Database Development
Satellite Data
Data Input Variables
2. Model
Development
3. Map Generation
Percent Annual Seasonal Greenness (PASG)
Start of Season Anomaly (SOSA)
Climate Data
Regression
Tree
Model (*)
Palmer Drought Severity Index (PDSI)
Standardized Precipitation Index (SPI)
Biophysical Data
1-km2 VegDRI Map
land use/land cover type
soil available water capacity (STATSGO)
ecoregion type
irrigation status
elevation
* Models developed from a 20-year historical record (1989 – 2009)
of climate and satellite observations at 3,000+ weather
station locations.
Biophysical variables are static over time.

19. VegDRI Flow: Providing Near-real time Delivery of Information
eMODIS Historical
Input Data Target: Monday 10:30 a.m.
Terra MODIS
T+11hrs
LAADS
USGS Drought Monitoring
VegDRI
MODAPS
EDOS
T+6hrs
NIDIS Drought Portal
MODIS L0 Data
T+3hrs
eMODIS Expedited
NDMC Vegetation Drought
Response Index
U.S. Drought Monitor
MODIS
L2, L1B Data
LANCE

20. 8-day Cumulative ETa Anomaly
Apr 1– Sep 12, 2012
Courtesy: G. Senay

21. 2001, 2006 National Land Cover Datasets
20 LC Classes
Only 2 Agricultural Classes
~78% Overall Accuracy

22. Information for decision-makers through combinations of
Information/data

23. Crop Type Classification of the U. S
Crop Type Classification of the U.S. Great Plains: An Application of Regression Tree Modeling using Remote Sensing and Environmental Data
Spatial coverage: U.S. Great Plains
Temporal coverage: 2000 – 2011
Spatial resolution: 250 meter
Overall model classification accuracy: 87%
Linear agreement with county results from the USDA NASS Survey Program: R2 = 0.76
D. Howard, B. Wylie

24. Global Agricultural Monitoring System of Systems:
GEOSS Activity
Satellite observations, in-situ observations and model outputs are needed.
Standardization and coordination among countries
Combined information (seasonal forecast models, agro-meterological data [precip, temp, humidity], in-situ obs, satellite obs at various scales, optical, thermal, microwave…
Planted area, crop condition, crop type, crop yield
To Build a Global Agricultural Monitoring System of Systems:
Satellite observations, in-situ observations and model outputs are needed to provide information in support of these themes. For example, a combination of seasonal forecast models, agro-meteorological data on rainfall, temperature and humidity, in-situ observations on rainfall and temperature, satellite observations at coarse, moderate and fine scales from optical, thermal and microwave sensors and sample field reports on area planted and crop condition are used to provide critical and timely information on crop type, area planted, crop growth and condition and crop yield.
For agricultural monitoring, multiple spatial and temporal scales of data are needed. In general, satellite observations at five broad spatial scales are needed to monitor agricultural lands. Global coverage of coarse resolution geostationary meteorological satellite data (5km-1km) are needed to provide hourly monitoring of weather conditions and rainfall. Daily, global coverage of coarse resolution polar orbiting data (1km -250m) are needed to provide a cropland mask and to monitor vegetation state and identify anomalies. At the national and sub national scale, two to three coarse to moderate resolution observations (250-20m data) are needed every ten days to provide crop type area and crop specific conditions and anomalies. One to two images of moderate resolution data (20-5m) are needed every ten days to provide information on crop type, crop stage and other crop variables at the parcel level. One to two fine resolution observations (5-1m) per month are needed to provide sample point interpretation and sub-parcel variability, which can feed crop growth models used in yield estimation.

25. Summary
Remotely sensed observations (spatial detail and synoptic coverage) are a critical tool for monitoring hazards
Need to understand limitations and accuracy
Need for continued development of tools that fit national needs, and also enhance a global system
International cooperation and technical advancement is being supported through GLAM

26. Thank You! Muchos Gracias!
Questions? Preguntas?

27. Extra slides

28. 2006 Satellite Vegetation Phenology for the Conterminous U.S.
April 2, 2006
April 30, 2006
May 28, 2006
June 25, 2006
July 23, 2006
August 20, 2006
September 17, 2006
October 15, 2006
October 29, 2006
U.S. Geological Survey
U.S. Department of the Interior
March 2007

29. Earth Observing Systems
VIIRS optical band widths will be similar to MODIS

30. Integrated Assessment
and Planning
Courtesy: D. Wilhite

31. Examples of Federal Policy Decisions
Allocation of Federal Emergency Relief Funds
U.S. Drought Monitor used as primary tool in allocating federal emergency drought relief
Bureau of Reclamation for American Indian Nations of the Southwest
Use of US Drought Monitor to trigger several key drought mitigation programs (including livestock assistance under the non-Fat Dry Milk Initiative and the Conservation Reserve Program)

32. Examples of Agribusiness Decisions
Federal Drought Information in Agribusiness
Used by Farmers and Ranchers in making purchasing decisions
Making crop planting decisions
Identifying areas of greatest demand for hay
Evaluating feed supplies and potential for shortages
Kansas Association of Wheat Growers and other agriculture-related groups rely on US Drought Monitor products

33. Examples of Agribusiness Decisions
Futures pricing of U.S. Commodities
Kansas City and Chicago Boards of Trade use federal drought monitoring and forecast information in futures pricing for U.S. Commodities
Other Futures brokering companies (e.g., Allendale Inc.) use the USDM for assessing impacts to grain futures and for providing advisories to customers

34. Examples of Water Management Decisions
Bonneville Power Administration with $3B in annual revenues depends on water supply forecasts/drought information for hydro power management decisions
Critical water management decisions that affect agriculture and wildlife resources in Klamath Basin are heavily dependent on accurate and timely drought information

35. Examples of Energy Decisions
Federal Energy Regulatory Commission
Use Drought Monitor information for the oversight of energy markets
Hold drought workshops to improve and maintain coordination and cooperation among licensees, agencies, stakeholders, and Commission staff during developing drought

36. State and Local Decisions
State and local drought planning
Requires higher spatial resolution and improved method for communicating drought classification categories
Enhanced coverage of surface observing networks along with improvements in reporting frequency and timeliness will support decision-making at the county level