Improvement and extension of IFM-CAP model: endogenous crop yield adjustments

About

Date: 2015-2016

Funding source: European Commission – Joint Research Centre (JRC)

Project title: Improvement and extension of IFM-CAP (Individual farm Model for the Common Agricultural Policy) model

Project summary

In the current context of resource depletion in intensive agriculture, IFM-CAP has pointed out the need of extending the current IFM-CAP model with yield response functions towards main inputs: water and fertilizer. The overall goal is to determine how farmers will adapt upon policy incentives by changing their main input use. In order to answer to this question, we create response curves to inorganic and organic fertilizer taking into account interactions between water, fertilizer and soil characteristics. Endogenous yield response functions will be applicable at EU level.
This report describes the different process steps in order to create and validate endogenous yield curves, and is organized in different section or tasks:

  • Task 1: Selection and justification of the methodological framework
    Background of main approaches in the economic literature to estimate yield response functions, as well as existing crop models that are used for this purpose are described at this step. The choice of the mechanistic EU-Rotate_N model is mainly justified by the importance to focus in soil dynamics and the computational and data requirements of the model.
  • Task 2: Identification of data needs and data quality evaluation
    Input data needs, data sources and main data assumptions are summarized in task 2. One of the contributions of this project is to point out the importance of soil characteristics in order to create accurate yield curves. To this goal, we choose to make yield simulations at a less aggregated level (HSMU levels) by using spatial units that are currently used in the CAPRI model. With respect to data needs of the crop model, site and soil data are mainly obtained at HSMU level from the European Soil database. An algorithm has been created to extract weather data from the NUTS-2 ascii raster dataset for EU-28. European Climate Assessment & Dataset has been used for this purpose. Finally, other data comes from JRC-MARS Crop Calendar dataset and FADN data.
  • Task 3: Data treatment and module implementation
    An automatic model in GAMS is proposed to be used as interface between the EU-Rotate_N crop model on the one hand and different datasets at HSMU and NUTS-2 levels on the other. The automatic program basically consists of the iterated execution of the EU-Rotate_N crop model at HSMU and crop level. Results will be later aggregated at NUTS-2 level within the R software, in the curve fitting step.
  • Task 4: Empirical implementation and validation
    Task 4 summarizes main findings of the project. We show that yield response functions are more affected by a change in nitrogen inorganic fertilizer level, than in organic fertilizer, as shown in statistical regression results for different fitted functional forms. We conclude that there is not a "best to fit" functional form for every crop and region. Each region needs a specific statistical analysis, to adapt the "best to fit" functional form for every crop. However, as reviewed in the literature on yield response functions and summarized in the validation step, a quadratic fit seems the most appropriated alternative to the current IFM-CAP model procedure. Moreover, yield response functions for pesticides are not feasible at this stage of the work because of the absence of practice information at farm level.

There are possible extensions to his work. The construction of the yield curve was based on spatial units (HSMU) used in the current CAPRI model. Another project on spatial allocation based on the definition of new spatial units at the European level is now running. More accurate new spatial units could be then incorporated to this work in a further project. Other possible extensions could consist of modelling the rest of crops presented in the crop model, taking into account crop rotations, as well as to make distinctions between winter and summer crops.