Data? Mike R. Duncan, Ph.D. Sarah Lepp, B.Sc. Gregor Maclean The - - PowerPoint PPT Presentation
How Much Data is Enough Data? Mike R. Duncan, Ph.D. Sarah Lepp, B.Sc. Gregor Maclean The 3D Vineyard Engine (2006) Creation of a 3D data-driven visualization model from 8pt/m2 LiDAR data. The vineyard engine was connected to a
Mike R. Duncan, Ph.D. Sarah Lepp, B.Sc. Gregor Maclean
Converted the 3D Vineyard Engine to Google Earth with 300,000 vines in the DB. Reynold’s GPS mapped Sentinel Vines.
0.5 1.0 1.5 2.0 2.5 3.0 3.5
Shows that the temperature field is as turbulent as Kolmogorov air turbulence. Characterized by sudden violent changes. Suggests a need for high frequency sampling to capture all the variability.
8 13 18 23 28 03/05/2012 0:00 04/05/2012 0:00 05/05/2012 0:00
Data from the Niagara College R-T Temperature Sensor Network – May 3rd, 2012
March 27th, 2012 and killed much of Ontario’s tender fruit crop.
1 3 5 7 9 26/03/2012 0:00 26/03/2012 12:00 27/03/2012 0:00 27/03/2012 12:00 28/03/2012 0:00
Temperature in deg C Data from the Niagara College R-T Temperature Sensor Network – March 27th, 2012
200 400 600 800 1,000 1,200 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 (million lbs)
Marketed Production of Fruit Crops, Ontario, 1985-2015
apples grapes peaches strawberries pears sour cherries sweet cherries plums & prunes raspberries Total
March 27 Killer cold event reduced production by >300 million lbs.
9
Yield Monitor Good GPS device to map Yield Monitor output Well tuned and well maintained harvester. Keep in mind that there’s a lag between when the harvester takes in yield and when the monitor sees it – this distance can be largish – so you need to correct for it. Vines/grapes are a very managed yield which makes it hard to verify the terrain/yield relationships. In 2009 we started looking at grains and grain farms. Opened the door to making map tools Yellow-Gold Farms in Parkhill Ontario.
LandMapR 4 Landform Map for Frasier Field in Parkhill Ontario 2002 Corn Yield map for Frasier field
surface data
0.005 0.01 0.015 0.02 0.025 0.03 0.035 20 40 60 80 100 120 140 160 Landform 1 Landform 2 Landform 3 Landform 4
0.005 0.01 0.015 0.02 50 100 150 200 250 300 Landform 1 Landform 2 Landform 3 Landform 4
0.05 0.1 0.15 0.2 0.25 0.3 0.35 1 2 3 4 Gaps (differences) in population between over and under performing cells 1 = P<20%, 2 = P<40%, 3 = P<70%, 4 = P=100% LandForm 1 LandForm 2 LandForm 3 LandForm 4
The correct way to compare landforms and crop performance is via a yield probability map. This graph shows that landform 1 shows a performance deficit throughout the full range of yield values. The yield probability index is a reclassification of yield values into performance values relative to the average performance of the crop for a year. The values cluster and the clusters closely follow the landforms
Local Watersheds Global Watersheds Wetness and Streamflow
maps with validation built-in!
Co-operator data submitted + collect geospatial data to fill gaps Goals: wireless data transfer & analyze data layers with transparent mathematics for teaching farmers Rx maps: implemented with industry direction, support
This project was funded in part through Growing Forward 2, a federal-provincial-territorial initiative. The Agricultural Adaptation Council assists in the delivery of Growing Forward 2 in Ontario.
Upload Cleaning Gridding/Mapping (Kriging) YI Elevation YPI/YPZones
YCI
EC Partitioning
Analytics Yield Elevation Soil Type
Chemistry Nutrients
Velocities, direction, boundaries, etc…
Inputs VRx RRx Variable Rate Sampling
Measurement Test Plots
Field Work
Elevation
Clipping the distribution to remove high and low
Function removes << 1% of the data most of the time. Not the best way to do business.
Y
V
A
dY dV
50 100 150 200 250 20.000 40.000 60.000 80.000 100.000 120.000 140.000 160.000 180.000 200.000 220.000 Yield in bu/ac Corn 1996 Corn 1998 Corn 2000 Corn 2002 Corn 2005 Corn 2007
50 100 150 200 250 300 350 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Histogram Count Normalized Yield Corn 1996 Corn 1998 Corn 2000 Corn 2002 Corn 2005 Corn 2007
Single trait seed Multi trait seed
0.1 0.2 0.3 0.4 0.5 0.6 0.7 1 2 3 4 5 6 7
Yield Stability Always Converges onto Areas of the Farm Field
Always Over-Performing Area Always Under-Performing Area
Percent of Field Area
Non-normalized distributions show an average yield that varies greatly from year to year
2000 4000 6000 8000 10000 12000 14000 16000 30 80 130 180 230 280 330 Yield in bu/ac Corn 2001 Corn 2002 Corn 2004 Corn 2005 Corn 2008 Corn 2010 Corn 2011
YPI Membership Distributions YPI Membership YPI Clusters Historical Yield Maps Avg Yield Cells (6m2) YPI Level
1000 2000 3000 4000 5000 6000 7000 110 130 150 170 190 210 Yield in bu/ac P=7 P=6 P=5 P=4 P=3 P=2 P=1 P=0
Green is High White is Low
2000 4000 6000 8000 10000 12000 110 130 150 170 190 210 Yield in bu/ac P=6&7 P=3&4&5 P=0&1&2
Green is High Pink is Low Yellow is Mid
Push Button Operation. Uses a yield map, or yield index map as a pattern and takes a target yield to generate removals. Fertilizers are added to match the removals. There are seven resulting maps.
But this area can’t be any good… it’s full
Yield Performance Index Map Green is High-Performing areas Red is Under-Performing areas.
A complex calculation involving historical yield and field topography performed by the NC Research Crop Portal. Another complex calculation performed by LandMapR using field topography, or elevation data. The figure shows local watersheds in the field, or where water will first pool under rain or irrigation inputs.
10 11 12 13 14 15 16 40 45 50 55 60 65 70 75 80 16/05/2016 20:09:36 16/05/2016 20:38:24 16/05/2016 21:07:12 16/05/2016 21:36:00
May 16, 2016 - 20:58(ish)
RH(Low) RH(High) T(Low) T(High)
No Prior warning of the sudden drop in either RH%
moving along and then a drop….
Humidity RH(%) Temperature deg C Time
10 20 30 40 50 60 70 80 25 30 35 40 45 50 1 299 597 895 1193 1491 1789 2087 2385 2683 2981 3279 3577 3875 4173 4471 4769 5067 5365 5663 5961 6259 6557 6855 7153 7451 7749 8047 8345 8643 8941 9239 9537 9835 10133 10431 10729 11027 11325 11623 11921 12219 12517 12815 13113 13411 13709 14007 14305 14603 14901 15199 15497 15795 16093 16391 16689 16987 17285 17583 17881 18179 18477 18775 19073 19371 19669 19967 20265 20563 20861 21159 21457 T(T) T(M) T(B) Therm(T) Therm(B) RH(T) RH(M) RH(B)
Nightfall – characterized rising humidity and falling temperatures, as well as inversions.
Temperature data points at 3cm and 1.2m mapped onto Google Earth. Data rate is 1Hz. Vehiccle velocity is ~3 mph