Skip to main content

User account menu

  • Log in
Home

Navigation principale

  • Home
  • About us
    • CLAND Convergence Institute
    • CLAND Partners
    • CLAND Newcomers
  • Research
    • Our research
    • Challenge 1
    • Challenge 2
    • Challenge 3
    • Publications
  • Graduate Program
    • Our Graduate Program
    • Our Internship Program
    • 2021 Call for Master Grants
    • Internship Offers
  • News
    • 2021 Call for PhD topics
    • 2020 Call for PhD topics
  • Events
    • 2021
      • Nature-based solutions
    • 2020
      • Machine Learning
      • Albedo & Climate...
      • General Assembly
      • Newcomer's Day
      • CLAND / SAPS
      • IPCC report & Students
    • 2019
      • IPCC Report on Land
      • 4 per 1000 initiative
      • ISIMIP II
      • ISIMIP
    • 2018
      • Forecasting Crop Yields
      • Soil Carbon Sequestration

Breadcrumb

  • Home
  • Forecasting Crop Yields from Data, Models, and Expert Knowledge

Forecasting Crop Yields from Data, Models, and Expert Knowledge


 

Date: 6 - 7 December 2018

Location: AgroParisTech premises, 19 Avenue du Maine, 75015 Paris


This workshop provided a unique opportunity for research scientists, engineers, and students to learn about recent crop yield forecasting techniques. Yield predictions are an essential source of information for a diversity of stakeholders. Policy makers or commodity traders can adjust their import/export plans and prices as a function of seasonal yield predictions. Farmers can directly benefit from crop yield forecasts by optimizing their management strategies, sometimes getting a better income. Yield predictions are also useful to help collecting firms planning their grain harvest and storage, and to support the decisions of companies investing in regional agricultural activities. Finally, yield forecasts can be incorporated in crop insurance systems to cover the risks of severe losses due to adverse climate conditions. Potential benefits obviously depend on the reliability of yield forecasting systems. The use of inaccurate predictions can lead to poor decisions and have negative socio-economic impacts. It is thus crucial to compare several forecasting systems and select the best ones. 

Crop yield forecasting systems rely on expert knowledge, remote sensing data, field observations, mechanistic and/or statistical models. Several of these techniques can be combined in order to improve the accuracy of forecasts. The recent development of machine learning methods offers new opportunities to synthetize various sources of information. A great diversity of methods is thus currently available, but the comparative advantages of recent methods compared to more traditional techniques are still poorly informed.

The workshop aimed at providing an overview of different types of yield forecasting methods and of their performances. The first part of the workshop was dedicated to plenary presentations of invited speakers, while the second part was dedicated to the presentation of the results of a data challenge on yield forecasting. Methodological contributions and practical applications were both welcome.

Université Paris-Saclay ANR

<3

 

Footer menu

  • Contact

@ Cland