14.2 Background
14.2.1 Related Work
Statistical offices around the world commonly use subjective procedures, similar to
NASS’s traditional method, to stratify area frames for agricultural surveys (Cotter
et al. 2010; Cotter and Tomczac 1994; Hanuschak and Morrisey 1977; FAO,
Volume 2 1998). The European Union’s Monitoring of Agriculture with Remote
Sensing project used visual interpretation techniques to conduct stratification for the
Land Use and Cover Area Frame Statistical Survey (LUCAS) and proposed using a
regular grid of square segments instead of physical boundaries for segment delineation to reduce cost (Gallego et al. 1994). The LUCAS sample units are points defined
as 3 m diameter circles (Gallego and Delincè 2010). The original objective of the
LUCAS was to provide annual European crop estimates. The Italian Agricultural
Survey (AGRIT) program uses an area frame created from a regular grid of points
with a resolution of 500 m. Point sampling is used for data collection for the
multipurpose survey to derive combined estimates of crop acreage, yields, and
land use (Carfagna and Gallego 2005; Benedetti et al. 2015). The Utilization du
territoire (TER-UTI) is an annual survey conducted by the French Ministry of
Agriculture, in which a grid area frame and sample points are used to collect land
use and land cover information as well as to derive an estimate of land cover change
(Benedetti et al. 2015). Workneh et al. used an area frame for evaluation of the
prevalence of brown stem rot in the north-central United States (Workneh et al.
1999). The European Union’s monitoring Agriculture through Remote Sensing
group also used an area frame for improvement of agricultural ground survey
estimates (Tsiligirides 1998). Faulkenberry and Garoui compared the utility of
frequently used estimators based on an area frame utilized in agricultural surveys
(Faulkenberry and Garoui 1991). Pradhan used an area frame for the development of
crop area estimation at a regional level in the Islamic Republic of Iran (Pradhan
2001).
Geospatial data such as aerial photography and satellite data are commonly used
to stratify area frames, delineate PSUs, delineate segments, identify point positions,
conduct data collection, and identify segment locations, using visual interpretation
and other manual techniques. Table 14.1 describes the types of sampling frames used
in a number of countries to conduct their agricultural surveys. The column on the far
right identifies how geospatial data are used in the area frame construction process
(FAO, Vol. 2 Report 1998). Martinez (2013) provides further details on the subjective use of geospatial data for area frame construction by countries around the world.
Over the past decade, quality geospatial data and technology have become widely
available at low or no cost to the public (USGS 2017c; ESA 2017a, b; QGIS, 2017;
Han et al., 2012). Utilizing remotely sensed geospatial land cover data for stratification of an area frame has been gradually investigated since 1990. The European
Coordination of Information on the Environment (CORINE) Land Cover 2000 map
was tested for use in stratification in Spain, but difficulties were encountered when
268
C. G. Boryan and Z. Yang
14.2.1 Related Work
Statistical offices around the world commonly use subjective procedures, similar to
NASS’s traditional method, to stratify area frames for agricultural surveys (Cotter
et al. 2010; Cotter and Tomczac 1994; Hanuschak and Morrisey 1977; FAO,
Volume 2 1998). The European Union’s Monitoring of Agriculture with Remote
Sensing project used visual interpretation techniques to conduct stratification for the
Land Use and Cover Area Frame Statistical Survey (LUCAS) and proposed using a
regular grid of square segments instead of physical boundaries for segment delineation to reduce cost (Gallego et al. 1994). The LUCAS sample units are points defined
as 3 m diameter circles (Gallego and Delincè 2010). The original objective of the
LUCAS was to provide annual European crop estimates. The Italian Agricultural
Survey (AGRIT) program uses an area frame created from a regular grid of points
with a resolution of 500 m. Point sampling is used for data collection for the
multipurpose survey to derive combined estimates of crop acreage, yields, and
land use (Carfagna and Gallego 2005; Benedetti et al. 2015). The Utilization du
territoire (TER-UTI) is an annual survey conducted by the French Ministry of
Agriculture, in which a grid area frame and sample points are used to collect land
use and land cover information as well as to derive an estimate of land cover change
(Benedetti et al. 2015). Workneh et al. used an area frame for evaluation of the
prevalence of brown stem rot in the north-central United States (Workneh et al.
1999). The European Union’s monitoring Agriculture through Remote Sensing
group also used an area frame for improvement of agricultural ground survey
estimates (Tsiligirides 1998). Faulkenberry and Garoui compared the utility of
frequently used estimators based on an area frame utilized in agricultural surveys
(Faulkenberry and Garoui 1991). Pradhan used an area frame for the development of
crop area estimation at a regional level in the Islamic Republic of Iran (Pradhan
2001).
Geospatial data such as aerial photography and satellite data are commonly used
to stratify area frames, delineate PSUs, delineate segments, identify point positions,
conduct data collection, and identify segment locations, using visual interpretation
and other manual techniques. Table 14.1 describes the types of sampling frames used
in a number of countries to conduct their agricultural surveys. The column on the far
right identifies how geospatial data are used in the area frame construction process
(FAO, Vol. 2 Report 1998). Martinez (2013) provides further details on the subjective use of geospatial data for area frame construction by countries around the world.
Over the past decade, quality geospatial data and technology have become widely
available at low or no cost to the public (USGS 2017c; ESA 2017a, b; QGIS, 2017;
Han et al., 2012). Utilizing remotely sensed geospatial land cover data for stratification of an area frame has been gradually investigated since 1990. The European
Coordination of Information on the Environment (CORINE) Land Cover 2000 map
was tested for use in stratification in Spain, but difficulties were encountered when
268
C. G. Boryan and Z. Yang