Bulletin on the National Cotton Output in 2016

National Bureau of Statistics of China

December 20, 2016

According to the result of nationwide statistical survey (the cotton sown area of Xinjiang was acquired through remote sensing survey) in 31 provinces (autonomous regions and municipalities) of China, the total cotton sown area, yield per unit area, and total cotton output in 2016 are as follows:

I. The total cotton sown area stood at 3,376.1 thousand hectares, down by 11.1 percent or a decrease of 420.5 thousand hectares compared with that in 2015.

II. The cotton yield per unit area amounted to 1,582.5 kg/ha, up by 7.2 percent or an increase of 106.5 kg/ha compared with that in 2015.

III. The total cotton output accounted for 5343 thousand tons, down by 4.6 percent or a decrease of 260 thousand tons compared with that in 2015.

Explanation on Cotton Output Survey

The national total cotton output was the sum of cotton outputs of 31 provinces (autonomous regions and municipalities).

1. Survey Methods

The statistics of cotton output adopted a combined statistical method, i.e, remote sensing survey and sample survey for the main producing areas, and comprehensive statistics for the non-main producing areas. Survey objects included land segments, agricultural production households and agricultural holdings. For Xinjiang, sown area data was acquired through remote sensing survey, while data of yield per unit area was acquired through sample survey. For Hebei, Jiangsu, Anhui, Shandong, Henan, Hubei, and Hunan, sown area data and data of yield per unit area were acquired through sample survey, carried out by the Survey Offices at provincial level of the NBS. For the other areas, sown area data and data of yield per unit area were acquired through comprehensive statistics, carried out by Statistics Bureau of provinces (autonomous regions and municipalities).

The agricultural production households were the objects of sample survey in the 7 main producing areas, and the survey was conducted by selecting representative villager groups, households and land segments, while taking a province as the population. Sown area survey referred to the check of total cotton sown area within the sample area at surveying time point, and the estimation of the actual acreage of cotton. Yield per unit area survey was conducted using field measurement method and family interviews, therefore, yield per unit area could be estimated. The total output data can be obtained by multiplying sown area and yield per unit area.

Starting in 2014, the NBS has conducted remote sensing survey of cotton in Xinjiang, which was carried out by combining medium and high resolution,  multi-temporal satellite remote sensing image, with ground-based sample survey, taking agricultural land segments as survey objects, then the region's cotton sown area on agricultural land was acquired by measuring. The sown area measured through cotton remote sensing survey and sample survey accounted for more than 95 percent of the total cotton sown area nationwide.

2.Survey Sample

The field investigation of cotton remote sensing survey carried out in Xinjiang included more than 2130 samples, or nearly 290 thousand mu of land segments. The 7 provinces conducted sample survey in 138 national survey counties, and more than 700 villager groups and 80 thousand households were selected in the cotton sown area survey. More than 700 villager groups and 7000 households were selected in the cotton yield of per unit area survey, conducted by staff members and assistant interviewers of survey offices at county level. In the other provinces (autonomous regions and municipalities) adopting comprehensive statistics, data was acquired through reporting level by level.

3. Field Measurement Method

According to The Program on Sample Survey of Agricultural Output, the survey of yield per unit area of main cotton varieties was conducted through field measurement and household survey.

Field measurement conducted a survey in random sample sites of the acreage survey. When the maturity stage of cotton approaching, the output of every piece of land growing cotton would be estimated, by means of systematic sampling grouped by yield per unit area, and ranked by accumulated acreage, 5 representative land segments were selected as field measurement plots. After the maturity of cotton, three small plots were selected artificially, and the yield per unit area of cotton was obtained by counting cotton bolls, weighting, estimating the losses, etc.

Household survey was carried out in random sample sites of the acreage survey. When the maturity stage of cotton approaching, the output of land segments of every household in the sample sites would be estimated, then the yield per unit area was obtained, and according to which, the average yield per unit area of each household was estimated. After the maturity of cotton, through systematic sampling which ranked by the average yield per unit area of each household and grouped by accumulated acreage, 10 households were selected, and after cotton picking, the cotton sown area, yield, sales, storing amount, own use amount of the households would be registered to calculate the yield per unit area. The Survey Offices at provincial level of NBS estimated the average yield per unit area of the province (autonomous regions and municipalities) according to yields of selected land segments.