To my knowledge there is no manual for crop yield estimation prior to harvest. But we take average of last five years and consider it average yield. Last contiguous five years includes both bad and good years. This average yield factor is used for multiplication with area covered in sowing for a crop. This is called estimated yield prior to harvest. Simultaneously we select 10 x 10 mt plots randomly at different locations and crop cutting/harvest yield is taken at maturity. This way average is taken and this crop cutting yield gives an estimation of actual yield after harvest and is used as factor for multiplication with area sown under particular crop.
Different approach is followed for crop yield forecasting depending upon the objective and/or the target area. Crop cutting, as Malhotra Sir has explained, is the most common method. However, other methods like thousand seed weight, farmers' estimate, whole plot harvest etc are also followed. Crop modelling is done for predicting yield in response to input management or climate variation. The following literature may provide useful clues to further review:
I was wondering whether from a managerial perspective it would be a wise idea to have different scenarios. Worst, average (something like what Dr. Borah and Dr. Malhorta were explaining) and some overly optimistic scenario. You want to have a plan for the worst situation, but you might also want to know if you can handle excessive quantities.
compliments to Christopher for critical remarks. These all are the methods given by statisticians. If you have some better method, please share with us, we shall welcome. What methodology you follow in your country for estimation of yield, we are interested to know. We have recently started using satellite based assessment system of area sown under particular crop/s and further crop yield models have been worked out. But for our colleague Mr. Esayas who hails from Ethiopia wanted an easy and feasible way to make estimation of yield of crop.
I read most of the crop estimation models are using satellite image or remote sensing technologies. I need a simple and easy method which can be used by farmers and development agents at county level in supply to the national estimation. thanks for the modules Dr. Nilay and suggestions of Suresh and christopher.