Agricultural surveys on area frame based on a squared grid
Area frame surveys have been used for a long time to obtain agricultural estimates including area and yield of major crops. Frames based on a square grid are much faster to define. A distance threshold can be used for sampling to avoid strongly correlated information from segments that are too close to each other. Forbidding pairs of segments at a distance less than d(0) modifies the sampling probabilities of the area close the border of the region when the threshold is applied on a simple random sampling, but leaves probabilities unchanged in an aligned sampling by repetition of a pattern in a squared block. When the sample is drawn by blocks, a cluster estimator can be defined as an alternative to standard formulae for random sampling. Setting up an area frame with a squared grid is much cheaper than with segments limited by physical elements. The precision of the estimates turns out to be very similar in both cases. The French TER-UTI survey reduced costs still further by visiting only a sample of points.
Bibliographic Reference: Paper presented: 50th Session of the ISI, Beijing (CN), August 21-29, 1995
Availability: Available from (1) as Paper EN 39239 ORA
Record Number: 199511233 / Last updated on: 1995-10-13
Original language: en
Available languages: en