|
Midwest | 1. If it's the same crop, then you don't have to since you're comparing apples to apples. (corn to corn) You still can if you like though.
2. I don't have any set number of years. If I have multiple crops (or a really bad year or two) I'll normalize it anytime I have 2 or more. Again, if it's the same crop then it's up to you.
3. There are mannnnny ways of normalizing the data. The ones I use the most are normalize on a 0-100% scale or normalize around data mean. 0-100% means it takes each year (dataset) individually and gives every dot or grid square a value from 0-100%. 3 years of data means you have 3 values from 0-100% for each square, and you get the average as your result. A really good spot might have 92 on year one, 94 year two, and 96 year 3. The normalized value would be 94 for that square. Another square could have 2, 85, and 66 for it's values and you'd get that average. All values will be between 0-100 and the higher the value the higher average.
If you're using the data mean option, that takes the mean of each dataset (year) and puts that at 100%. Anything over for that year is over 100% and anything under is under 100. Those are then averaged. 100 would then be your center of the data with a good spot reading maybe 136 and a poor area being 70.
The help text in SMS should give you some insight, but that's what I can recall off hand. (I'm in an ESRI class right now and they only have Arc on their computers, surprise surprise) | |
|
normalization of data
