Transcript CropMetrics_Holos2016_v3x

Crop Metrics
Building an Online System for Crop Information
Rolfe Antonowitsch and Crop Metrics Team
AAFC-Science and Technology Branch
1
Why Crop Metrics
• AAFC and partner organizations (Statistics Canada,
Environment Canada, provincial agencies) produce a
number of data sets that all provide crucial information
on the status and potential of crops during a growing
season
• A Crop Metrics application would provide a mechanism
to view this key information in a single report, using the
latest data sets and incorporating the best science
available.
2
Who are the clients for Crop Metrics?
1. Policy and Programs groups within AAFC-Risks
and response planning.
2. Markets Industry Services Branch (MISB)-Timely
reporting of Canadian crop inventories.
3. Provinces
4. Commodity Brokers and Grain Transportersadvanced planning/competetiveness.
5. Producer Groups/Producers-Information for inseason decisions/plans.
6. Research and other users- web enabled products
will enhance further reanalysis and use in other
products.
3
What Data Sets Can Support Crop Metrics?
4
• Main Inputs: NDVI and climate indices.
• Reporting time: Middle of July, August
September and October.
• Current forecasted crops: spring wheat,
durum wheat, canola, barley, oats, corn
for grain and soybean.
Empirical Cumulative density function
Canadian Crop Yield Forecasts
Best 90
Percentile
Median
Worst 10
Percentile
Forecasted Yield (kg/ha)
5
Current Data: Climate Data
•
•
•
•
Around 900 stations cross Canada
Around 450 stations in crop lands
Distributed unevenly
Basic inputs : Tmax, Tmin and precipitation
6
Growing Degree Day
Crop Accumulated Heat Metrics (July 15, 2015)
CAR 4841
Date
7
Accumulative Precipitation (mm)
Accumulated Precipitation Metrics (July 15, 2015)
CAR 4841
Date
8
Stress Index Metrics (July 15, 2015)
CAR 4841
Stress Index
Extreme Stress
Severe Stress
Light Stress
No Stress
Date
9
Current Data Sets: Satellite Crop Condition
Normalized
Difference
Vegetation Index
(NDVI) anomalies
show areas where
vegetation is less
vibrant than
compared to the
historical average
Anomalies can
occur because of
differences in
timing of seeding,
vegetation stress,
large scale changes
in crop mixture
10
• AAFC produces this data weekly through the growing season using the MODIS satellite
• Data are delivered through Statistics Canada Crop Condition Assessment Program (CCAP)
NDVI
NDVI Metrics (July 15, 2015)
CAR 4841
Date
11
Current Data Sets: Satellite Soil Moisture
Weekly, biweekly and
monthly soil
moisture and soil
moisture difference
from long term trend
(anomalies)
Sensitive to surface
soil moisture layer,
but gives an
indication of overall
moisture availability
Persistence of excess
wetness or dryness
at the surface can be
an indicator of
subsurface
conditions
12
Current Data Sets: Risks
North American Drought Monitor
•
Monthly Assessment of Drought Severity
Climate Production Risk Committee Reports
•
Summarize information nationally on key risks to
agriculture in each region
Agroclimate Impact Reporter (AIR)
•
Maps based on surveys of volunteer reporters
• Livestock feed shortages, excess moisture,
impact of below normal temperatures
13
Current Data Sets: Weather Forecasts (sample from August
2014)
14
Current Data Sets: Market Information
Chicago Board of
Trade Wheat
Futures
15
Potential Data Sets: Satellite Evaporative Stress
Legend
Evaporative Stress Index (12 weeks)
3 High Stress
-3 Low Stress
• Weekly measurement using thermal satellite data that shows areas were plant
water demand is higher than what is available (red areas)
• Produced by the US Department of Agriculture at 10km spatial resolution
16
Potential Data Sets: Pest Forecasts
Computer Centre for Agricultural Pest Forecasting
(CIPRA)
• Models risk of pests
and disease based on
climate station
information for a
number of key crops
in Canada
17
What Would a Crop Metrics Application Provide?
Integration of information that can be used for decision support by
regional experts
18
Crop Metrics Application
Crop Metrics Application
Select a Crop Type
Spring Wheat
Canola
Barley
Corn
Soybean
Select Point or Draw Area of Interest
Metric
Imperial
- Still in development, the Crop Metrics Application will make use of AAFC online
mapping capabilities to allow users to select and area and a commodity type to
explore
- Coloured map shoes area with the highest probability of finding the selected crop (in
this case, spring wheat)
19
Crop Metrics Report for AOI
Crop Metrics Application
Forecast Yield Potential
Worst Case: P10: 2.3 tonnes/hectare
Most likely: P50: 2.8 tonnes/hectare
Best Case: P90: 3.4 tonnes/hectare
Estimated Current Conditions (to current date)
Accumulated Heat Units : Favourable
Accumulated Precipitation : Poor
Satellite Vegetation Condition– Poor
Climate Stress Index – Severe Stress
Growth Stage – Ahead
Forecast Conditions
Precipitation: 50% Probability of Precipitation
Temperature: Above Normal Temperatures
Select a Crop Type
Spring Wheat
Canola
Barley
Corn
Soybean
Risks
Risk Factor 1 – Extreme Drought, worsening
Risk Factor 2 – Record Low Precipitation
Risk Factor 3 – Localized water shortages,
Below Average Feed Production
Select Point
Draw Area of Interest (AOI) or use
Use Predetermined Areas (ie townships)
This is still a work in progress – what
information is most important to
include?
Market
Commodity Price:
High: $328.55/tonne
Average: $288.77 /tonne
Low: $224.00 /tonne
Overall Outlook: Below Average
20
Crop Yield Metrics (July 15, 2015)
CAR 4841
Last 5-year mean
Yield (Bu/Ac)
Long term mean
July 2015 forecasted
P90, P50 and P10
Year
21
Our Next Steps
1. Calibrating Crop Metrics typology with clients
(starting with internal clients).
2. Exploring alternative methods od delivering the
metrics products apart from the web (Canada.ca
portal).
3. Integrating/confirming data sources from satellite
platforms, crow sourced avenues, GPS, pest
forecasts etc.
4. Developing synergies with other projects with
similar/related objectives (e.g. sustainability
metrics).
5. Present proof of concept to AAFC management in
Year 3 to decide on operational details.
22