Derniers produits inserés :
Rechercher un produit
Dernières contributions :
Télécharger le pdf "Article EJA 2014 CIRAD" , Paper "Coupling a sugarcane crop model with the remotely sensed time series of fIPAR to optimise the yield estimation” published by CIRAD in the n° 61 issue (...)

Télécharger le pdf "Article RS 2014 CIRAD" , Paper "Toward a satellite-based system of sugarcane yield estimation and forecasting in smallholder farming conditions: a case study on Reunion Island” (...)

Toutes les contibutions

Applications

Photo_Applications_Introduction

Context

Extracting agricultural information (yield, risk of diseases, stress...) directly from satellite images turned out quite disappointing until now. On the other hand, recent development of digital agronomic models makes it possible to simulate today farming practices and estimate their impact on production.

These models are based on many data (soil type, agro-meteorological conditions, crop treatments, irrigation, etc.) to predict crop status. Satellite imagery makes it possible to compare the crop forecast status with a concrete physical measurement.

Activities carried out by scientific teams

The assimilation technique, already used in an operational way in the fields of space oceanography or meteorology, therefore allows to adapt the models to the phenomena actually observed on the images and calibrate them on a regular basis.

The implemented methodology is an original one, as it is quite different from those considered in Earth observation until now. The issue is no more to extract information more or less directly from optical or radar satellite imagery, but to force the digital agronomic model describing the development of the studied crop to follow the spatial measurements regularly supplied through these images. The adjusted model has then to produce the requested information.

Conversely, integrating spatial measurements into models gives to space field researchers the opportunity to attract a new and quite sizeable user community. This community, linked with precision farming, is significantly expanding, considering the new environmental constraints and the necessity for the agricultural world to take them into account. Earth observation could therefore play a part in this expansion, once it will be demonstrated that the produced information are quite relevant for this advanced topic.

Relevance of a reference remote sensing database

To complete the Adam project successfully, it was necessary to collect a significant amount of spatial data, more especially high temporal repetitiveness Spot or radar images, and set up the appropriate research and analysis resources in the concerned laboratories and institutes.

A wide experimental operation has also been organised, more particularly during the 2000/2001 crop growing season, with a massive capture of meteorological and atmospheric data and the collection of field measurements (soil and crop characteristics, farming practices) carried out on 42 experimental units, each of them a little larger than a Spot pixel (diameter around 30 m).

The development of the Kalideos Adam database made it possible to structure this data set and provide researchers with a perennial frame to consult and use the included data.