CALIBRATION AND ATMOSPHERIC CORRECTION OF THE AVHRR 
FOR THE PURPOSES OF AGRICULTURAL MONITORING IN EUROPE 
M. Sharman 
Agriculture Project 
Institute for Remote Sensing Applications 
Joint Research Centre 
Ispra, Italy 
R. Santer 
Maître de Conferences 
Laboratoire d'Optique Atmosphérique 
Université des Sciences et Techniques de Lille 
Villeneuve d'Ascq, France 
A. Le Lerre 
Software Design and Implementation 
Remote Sensing Division 
Tecnodata Italia 
Frascati, Italy 
ISPRS Commission Number VII 
ABSTRACT 
The Commission 
established 
Communities has 
Remote Sensing 
(channels 
launch. 
radi a tion 
of the European 
Pilot Project of 
applied to Agricultural Statistics to monitor 
crop acreages and yields in Europe. This project 
uses high (SPOT and TM) and low-resolution 
(AVHRR) satellite data as sources of information. 
Although the high resolution data are 
satisfactorily calibrated, the visible and near 
infrared channels of the low resolution data 
1 and 2) are not calibrated after 
The absorption and scattering of 
by atmospheric constituents are not 
accounted for in the data as they are received at 
the earth stations. The project's requirements 
for calibration and atmospheric correction for 
the AVHRR are described and the means by which it 
will satisfy these requirements are outlined. 
The methods used for calibration include those 
described in Kaufman and Holben, consisting of 
calibrating the visible channel (Ch1) by taking 
advantage of the characteristics of light 
received by the sensor from a target over the 
ocean, inter-calibrating visible and near 
infrared channels using Fresnel reflection near 
sun glint, and using desert reflection for both 
C h1 and C h 2 calibration. Atmospheric correction 
is based on the "5S" model of Tanré e t a l . . but 
uses look-up tables to speed up processing. 
Key Words: 
AVHRR, Calibration, Atmospheric correction, 
Agricultural Monitoring 
I NTRODUCTION 
a. The Pilot Project 
One of the major underpinnings of the European 
Community is the Common Agricultural Policy. The 
depends on a complex system of rules and 
subsidies, which the Community now has to 
maintain and enforce. This system was set up in 
part because of the considerable differences in 
agricultural production between the various 
European nations. It relies for its operation 
and management on accurate and recent 
agricultural statistics. Unfortunately each 
European country monitors and measures its 
agricultural production in its own way, so that 
the basic statistics are rarely directly 
comparable, and the manager's task is one of 
great subtlety and skill. 
If the system is to 
manage, the various 
difficult to 
collecting, 
become 11 
systems 
analysing and reporting agricultural statistics 
must be upgraded and, especially, unified. 
Remote sensing seems to be the most promising 
technique. 
With these considerations in mind, the Commission 
of the European Communities therefore 
established, in September 1988, the Pilot 
Project of Remote Sensing applied to Agricultural 
Statistics. The aim of this project is to 
introduce remote sensing into the agricultural 
statistics system of the European Community as an 
additional source of data to complement the 
conventional existing sources. The Joint 
Research Centre is in charge of the project, 
whose orientation and aims are defined by the 
Directorate General VI (Agriculture) and the 
Statistical Office of the European Communities 
(OSCE ) . 
The results of this project will form part of an 
"Advanced System of Information on Agriculture". 
This system will be founded on remote sensing 
data from high- and low-resolution sensors and 
agro-meteorological models. In the pilot phase 
the emphasis is placed on remote sensing and 
agro-meteorological modelling, but it must be 
stressed that in operation, remote sensing will 
complement classical data, and is not expected to 
supplant them. 
The programme of the project includes 4 main 
actions: Action 1 (Regional Inventories) and 
Action 4 (Rapid Estimates at a European Scale) 
use high resolution imagery; Action 3 (Agro- 
meteorological modelling) uses mainly 
meteorological and soil data; while Action 2 
(Vegetation Condition and Yield Indices), the 
subject of this presentation, uses low-resolution 
AVHRR data . 
b. Action 2: crop monitoring with low resolution 
radiometers 
In its action 2, the project is to develop and 
demonstrate the use of low-resolution satellite 
data from the Advanced Very High Resolution 
Radiometer (AVHRR) on board the NOAA satellites 
for monitoring crop condition and in estimating 
178