# ABSOLUT Model ### Description [ABSOLUT v1.0 is a linear regression modeling algorithm](https://gmd.copernicus.org/preprints/gmd-2021-21/) that uses correlations between time-aggregated weather data and crop yields for yield predictions ([Conradt 2022](https://doi.org/10.1007/s00484-022-02356-5)). Within the DIRECTED project, ABSOLUT was used to produce crop yield scenarios, which can be found in the Rhine-Erft Region and the Danube Region RWL. The crop yield scenarios can be found in a button on the bottom-right labeled “Crop Chart.” ### Model Specifications | Model Specifications | Description | | ----------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Model Type | A statistically based approach utilizing multiple linear regressions. In contrast to other crop yield models of that type ABSOLUT self-optimizes its input meteorological regressor variables to sets with maximum predictive power. | | Associated Probability Method |

Standard errors of single results are among the outputs. Furthermore, the ten realizations used in each climate scenario create

empirical uncertainty distributions indicating the climate-model uncertainty.

| | Spatial Resolution | Principally any regional subdivision of the modelling domain for which observed weather and crop yield data exist. In the DIRECTED application these are the [NUTS-2 regions of the year 2021](https://ec.europa.eu/eurostat/web/products-manuals-and-guidelines/-/KS-GQ-22-010). | | Intensity Parameters | Regionally averaged crop yields of about a dozen staple crops in t/ha. | | Other Inputs | Monthly weather data. For the DANUBE application, these were spatially weighted monthly averages of E-OBS data. The spatial weighting was applied to each of the NUTS-2 regions considering the spatial distribution of agricultural areas. | | Calibration and Validation | The model is self-calibrating and implicitly considers systematic regional yield deviations, e.g. caused by soil quality or local management practices. Validation outputs (R-squared, RMSE, etc.) have been checked for hindcasting experiments but are not included into the Data Fabric. It may be noted that ABSOLUT could forward-predict German green maize yields over the years 2017–2024 with about the same fidelity (R-squared = 0.9, RMSE = 19t/ha) as the European Commission's forecasting model. In addition, ABSOLUT predictions can be made already in the first September week (as soon as the August weather data become available) whereas the EC predictions used for comparison were released in the fourth week. Finally, ABSOLUT works with weather data only whereas the EC forecast employs also satellite imaging and ground observations. | | Effects of Climate Change | These are explicitly modeled in this ABSOLUT application. The crop yield scenarios are based on the ISIMIP bias-adjusted climate scenarios for SSP 370 (modereate to high) and 585 (high emissions), consisting of ten realizations each. | --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://directed-eu.gitbook.io/data-fabric-manual/model-documentation/absolut-model.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.