It’s that time of year. The excitement is building. The patriotic bunting is being affixed to Cornbelt porches. High school marching bands are practicing. And on the morning of August 12, USDA statisticians will announce their forecast for the size of the 2010 corn and soybean crops. Yes, the August 1 Crop Report is just a few days away. And right now across the Cornbelt crop enumerators are accumulating a bumper crop load of data. But how do they do that, really?
The biggest crop report of the season began a few days ago when hundreds of crop enumerators fanned out across the Cornbelt to wade through corn and bean fields to collect data in objective yield surveys that is turned into statisticians at USDA to compile state and national yield projections. Additionally, they will interview as many as 27,000 farmers to get their estimation of the new crop yield. The process is intense, thorough, and quite scientific, if you read USDA’s Yield Forecasting Program.
Since the objective yield survey of 2,090 samples is a check against the survey of farmers, and vice versa, there is also a geographic check against a list of producers. A list frame is a roster of 50,000 producers with data on typical production. That data is compared to an area frame, which is an identified land area where crops are produced and detailed surveys are taken, by enumerators for the National Agricultural Statistics Service (NASS.)
For the August 1 Crop report, the survey process began about July 25 and will end on August 3. Enumerators have already picked fields at random and selected two random locations in each field to take crop samples. They will return to the same spots prior to the September, October and November surveys to make adjustments to their estimates. Just before the field is harvested the enumerators will harvest the sample and send it to laboratories for weight and moisture tests, ahead of the Final Crop Report in January.
Despite the scientific nature of the sampling and the complex calculus formulas developed over time in an effort to achieve the greatest possible accuracy, there will be biases. The NASS statisticians know that bias is a constant, and thus can be corrected. For example, if farmers tend to underestimate their yield by 2.3%, then the data from farmer surveys will be adjusted upward by that amount.
For corn the NASS enumerators are sampling 2,090 plots in the ten major corn producing states. NASS indicates, “A sample consists of two independently located units (or plots), each of which consists of two parallel 15 foot sections of row. Field enumerators use a random number of rows along the edge of the field and a random number of paces into the field to locate each unit. At each visit, enumerators count all fruit and fruiting positions.” Ears will be counted, and will be mailed to a laboratory just before harvest, and following harvest, the enumerators will return to the location and determine the harvest loss, in an effort to compute what was harvested.
Enumerators will compute the distance between the two rows, as well as the distances separating five rows. They will count the stalks, stalks with ears, and ears with kernel formation. They will measure kernel row length from the first five ears, beyond the sample row, the ear diameter, weight of ears in the dent stage, weight of shelled corn, and field weight of all ears in the two 15 foot long units. If their random rows and paces take them to the best spot in the field, a good yield will be recorded. If their random rows and paces take them to one of the drowned spots in a field, the yield will be recorded as a zero.
In some Cornbelt fields, the crop is just pollinating and is far from being able to be shelled for weighing. In those cases “All unharvested samples in maturities 1 and 2 (no ear shoots and pre-blister stage) use a 5-year historical average ear weight. Maturities 3-6 (blister through dent stage) employ models derived using kernel row length and ear diameter. Enumerator harvested samples use the average field weight per ear, shelled grain weight per ear, and the moisture content to estimate the grain weight per ear at 15.5 percent moisture.”
For the August report, the forecast yield will utilize ear and kernel information in IL, IN, IA, and NE. Only stalk information will be used in MN, OH, and WI. For September, ear and kernel information will be used for all seven of those states. Final data will be tallied also for MO, SD, and KS.
Tomorrow: soybeans
It’s that time of year. The excitement is building. The patriotic bunting is being affixed to Cornbelt porches. High school marching bands are practicing. And on the morning of August 12, USDA statisticians will announce their forecast for the size of the 2010 corn and soybean crops. Yes, the August 1 Crop Report is just a few days away. And right now across the Cornbelt crop enumerators are accumulating a bumper crop load of data. But how do they do that, really?
The biggest crop report of the season began a few days ago when hundreds of crop enumerators fanned out across the Cornbelt to wade through corn and bean fields to collect data in objective yield surveys that is turned into statisticians at USDA to compile state and national yield projections. Additionally, they will interview as many as 27,000 farmers to get their estimation of the new crop yield. The process is intense, thorough, and quite scientific, if you read USDA’s Yield Forecasting Program.
Since the objective yield survey of 2,090 samples is a check against the survey of farmers, and vice versa, there is also a geographic check against a list of producers. A list frame is a roster of 50,000 producers with data on typical production. That data is compared to an area frame, which is an identified land area where crops are produced and detailed surveys are taken, by enumerators for the National Agricultural Statistics Service (NASS.)
For the August 1 Crop report, the survey process began about July 25 and will end on August 3. Enumerators have already picked fields at random and selected two random locations in each field to take crop samples. They will return to the same spots prior to the September, October and November surveys to make adjustments to their estimates. Just before the field is harvested the enumerators will harvest the sample and send it to laboratories for weight and moisture tests, ahead of the Final Crop Report in January.
Despite the scientific nature of the sampling and the complex calculus formulas developed over time in an effort to achieve the greatest possible accuracy, there will be biases. The NASS statisticians know that bias is a constant, and thus can be corrected. For example, if farmers tend to underestimate their yield by 2.3%, then the data from farmer surveys will be adjusted upward by that amount.
For corn the NASS enumerators are sampling 2,090 plots in the ten major corn producing states. NASS indicates, “A sample consists of two independently located units (or plots), each of which consists of two parallel 15 foot sections of row. Field enumerators use a random number of rows along the edge of the field and a random number of paces into the field to locate each unit. At each visit, enumerators count all fruit and fruiting positions.” Ears will be counted, and will be mailed to a laboratory just before harvest, and following harvest, the enumerators will return to the location and determine the harvest loss, in an effort to compute what was harvested.
Enumerators will compute the distance between the two rows, as well as the distances separating five rows. They will count the stalks, stalks with ears, and ears with kernel formation. They will measure kernel row length from the first five ears, beyond the sample row, the ear diameter, weight of ears in the dent stage, weight of shelled corn, and field weight of all ears in the two 15 foot long units. If their random rows and paces take them to the best spot in the field, a good yield will be recorded. If their random rows and paces take them to one of the drowned spots in a field, the yield will be recorded as a zero.
In some Cornbelt fields, the crop is just pollinating and is far from being able to be shelled for weighing. In those cases “All unharvested samples in maturities 1 and 2 (no ear shoots and pre-blister stage) use a 5-year historical average ear weight. Maturities 3-6 (blister through dent stage) employ models derived using kernel row length and ear diameter. Enumerator harvested samples use the average field weight per ear, shelled grain weight per ear, and the moisture content to estimate the grain weight per ear at 15.5 percent moisture.”
For the August report, the forecast yield will utilize ear and kernel information in IL, IN, IA, and NE. Only stalk information will be used in MN, OH, and WI. For September, ear and kernel information will be used for all seven of those states. Final data will be tallied also for MO, SD, and KS.
Summary:
Statisticians from the National Agricultural Statistics Service are currently working across the Cornbelt, taking samples of corn, in an effort to make a yield projection for the August 1 Crop Report. Nearly 2,100 samples are being measured, both in the field and on the cob to arrive at a statistically accurate projection. Double checks are being made, using both crop enumerators and farmers, as well as types of surveys to ensure there is a minimal amount of bias in the report.
Statisticians from the National Agricultural Statistics Service are currently working across the Cornbelt, taking samples of corn, in an effort to make a yield projection for the August 1 Crop Report. Nearly 2,100 samples are being measured, both in the field and on the cob to arrive at a statistically accurate projection. Double checks are being made, using both crop enumerators and farmers, as well as types of surveys to ensure there is a minimal amount of bias in the report.