58 
REMOTE SENSING AND GIS FOR AGRICULTURAL CROP ACREAGE 
AND YIELD ESTIMATION 
Vinay K. Dadhwal 
Space Applications Centre, Ahmedabad 
ABSTRACT 
The use of satellite remote sensing data (RS) for operational crop assessment and forecasting at regional scales is of great significance for 
food security and policy decisions. Improvements in sensor capability and analysis techniques now allow accurate crop discrimination. 
Through use of GIS, regional sampling strategy using land cover, crop distribution and administrative boundary information, large area 
crop inventory is carried out. The use of RS data for yield forecasting, with high and coarse resolution and single or multiple acquisitions 
has generally followed empirical regression approach. The information on soil and weather can be integrated for yield modelling through 
use of crop simulation models. GIS plays an important role in converting the input data to a common format for use in the model. The 
recent Indian experience in developing techniques for estimation of 
summarised. 
REMOTE SENSING IN CROP ACREAGE 
ESTIMATION 
The use of spaceborne RS data for large area crop 
survey was first explored under CITARS Project in USA 
and was followed by an attempt to forecast wheat for 
major growing regions of the world under LACIE 
(Large Area Crop Inventory Experiment, 1974-1977) 
(MacDonald and Hall, 1980). Since then, large scale 
methodology development-cum-demonstration studies 
for crop statistics have been carried out in Africa and 
Europe as well as in a number of other countries 
(Argentina, Australia, Brazil, Canada, Japan etc.). 
Currently major programs are underway in Europe under 
MARS (Monitoring Agriculture through Remote 
Sensing). 
Remote Sensing (RS) can be used for crop acreage 
estimation in either for making area sampling frame 
(crop inventory being done by field survey), or for crop 
discrimination. Current regional crop inventory studies 
use RS for both applications, GIS-based sampling and 
digital image processing for crop discrimination. In case, 
the accuracy of RS-based discrimination is not 
acceptable, it has to be combined with field survey 
information. 
RS-BASED ACREAGE : INDIAN EXPERIENCE 
CAPE PROJECT 
Systematic studies on district-level crop inventory 
were taken up for single crop dominated districts 
(Karnal, Haryana for wheat, Dadhwal and Parihar, 1985) 
and promising results led to state level wheat acreage 
estimation in Haryana and Punjab using a sample 
crop area and forecast of likely yield using both RS and GIS is 
segment approach. It used 10 x 10 km segments, a 
stratified sample design and 10 percent sampling 
fraction and a single acquisition at optimal bio-window 
(Dadhwal, 1986). This approach has been followed in 
Crop Acreage and Production Estimation (CAPE) Project, 
which is sponsored by Ministry of Agriculture and covered 
six major crops in 15 States (Navalgund et al., 1991). The 
study areas for these crops are summarised in Table 1. A 
summary of results of CAPE project for 1997-98 crop 
season are provided in Table 2. A procedure using multi 
date WiFS data and GIS for sample design, developed for 
national-level multiple wheat assessments (Oza et al, 
1996) is described below. 
Improving the accuracy of RS-based crop inventory 
The CAPE procedure is being continuously revised 
and upgraded to improve upon accuracy and timeliness of 
crop estimates. These efforts are related to improving a) 
Sample design, b) Ground truth data collection, c) 
Optimising date of data acquisition, d) Including data from 
additional spectral regions in the digital analysis, e) Multi 
date data analysis, f) Use of higher spatial resolution data, 
g) Adopting different classification procedures, and h) Use 
of microwave data for crop inventory in kharif season. 
Significant results from these studies are summarised 
below. 
Sample Design 
Studies carried out on rice for Orissa and wheat for 
Haryana have clearly shown that two-step stratification 
based on agro-physical and crop distribution improves 
efficiency of stratification (Panigrahy et al., 1991). Size of 
the sample segments and sampling fraction are two