AgTalk Home
AgTalk Home
Search Forums | Classifieds (60) | Skins | Language
You are logged in as a guest. ( logon | register )

Climate Field View vs JD Operation Center
View previous thread :: View next thread
   Forums List -> Crop TalkMessage format
 
GOOSEPILOT
Posted 3/21/2020 09:48 (#8125184 - in reply to #8122705)
Subject: RE: Climate Field View vs JD Operation Center


WC Mn/Dakotas
mnfarmboy - 3/20/2020 14:19

From what I know, JD does not offer vegetation maps like FV does. Also, prescriptions on FV can be drawn with Any background map and can be switched while drawing rxs. Jd does not. Neither can let you average hand picked maps to make a Rx


These images can be manipulated 1000s of different mathematical ways. You need to know what your trying to identify and what script will so that. Is the image bare soil, low vegetation, dense vegetation, actively growing, maturing, etc. Images really all need to be purposely selected and handled appropriately.

Then these images really need to be slightly georeferenced specifically to each field. They all vary. I go through each image, they almost all need to be shifted in 1 or 2 different directions as each one "floats around" 10-200 feet. Very few images dont need adjustments after cutting down to the field level.

If you dont do this it isnt necessarily wrong, it's just not going to be nearly as accurate as it could be. I know the way I handle my images and the way I build my zones, my in field accuracy can identify spots as small as 10 yards in diameter and do it within inches of reality in the field. If you didn't geo adjust each image, the resulting zone might make that 10ft diameter look like it's 50 ft in diameter and infield accuracy could be of 10 feet or more. Some of our zones (ditches or knobs) are only 10ft total.

I used this as an example that easy easy to ground truth and measure results. I realize 10ft agruable isnt worth messing with. But illustrates the level of accuracy every time a rate change is commanded 100% of the time. I would argue though that it is worth adjusting for. If I have Elec meters on a planter, I can put 16k-20k seeds on that poor zone and in a few feet be back up to 28-32k in the good zones. Or here we don't lime every acre, it might just be those "knobs" are 6.0 Ph and a few feet away the ph is 7.5

Why would someone want to "look" at a single image (or whatever layer) and hand draw a zone? Is that very precise? Now with sentinel imagery we get an image every week basically. I might use 5-12 images from a single year to represent the year accurately. Then i usually use at least 10 years of imagery (not to mention the other forms of data layers) to build my zones. So im making my zones off of 30-50 images. We have computers and there are in depth softwares that can process and "draw" a whole lot better than my finger can.

Then yield data. Sure you can use raw data, but again even the best calibrated and most carefully operated machines are going to produce some bad data points that if not cleaned up, skew the results.

In GK I will take my zone layer and summarize the raw yield by each zone. You get an answer of what the min, max, and avg yield is in each zone. Then i run the data through Yield Editor and summarize that by the zones again. Now it gives you a different answer. Then i further manually clean up the yield editor version in gk and summarize that version.

Some zones the avg yield changes 10-15 bushels (in soybeans thats huge) or the min and max range is reduced 50-75 bushels (in corn) that's way more accurate.

This is more of a response to my other post in this thread about decisions you want to make from this data.

I think many growers dont understand what goes on behind the software and just assume the easy answers are all simple and accurate. They are not simple and easy. If they are, they are not very accurate if you dig into them at all.

I personally would not use either myjd or ccfv as a grower to decide anything other than what areas to spend time doing a thorough and accurate analysis of with some dedicated analytical software. They are great for monitoring equipment and personnel, for taking data with you to ground truth in the field, or visit with consultants, or do a quick casual look back at field history.

Making decisions based or bad or poor data is dangerous and can be extremely costly. Having a false confidence in making these decisions on bad or poor data is even worse. Just be careful understand what your looking at.
Top of the page Bottom of the page


Jump to forum :
Search this forum
Printer friendly version
E-mail a link to this thread

(Delete cookies)