Review of crop modelling approaches to address climate change challenges in Africa

. Africa is facing an urgent need to increase food production to meet increasing demands. Targeted investments in integrated agriculture and, water management systems are required to meet this challenge. However, there is a lack of comprehensive information on the potential applications of climate-smart agriculture (CSA). This paper reviews current crop modeling technologies and their applications within the scope of climate change and the CSA framework in Africa. It evaluates current research trends in various crop simulation models and suggest advanced approaches to improve crop and environmental assessment, crop management, and decision-making. A total of 140 relevant papers were considered. Results showed that 84% of studies used process-based models, with Maize being the most studied crop. Additionally, DSSAT crop models and analysis of variance models have the highest contribution of physical and empirical crop modeling studies respectively. Over 72% of studies have contributed to adaptation strategies and reducing yield gaps, while only 8% of studies have been conducted on climate change mitigation and their trade-offs with adaptation using crop models under CSA. To ensure food security through sustainable agricultural practices in Africa, there is crucial to implement CSA models with a focus on the climate change mitigation component.


Introduction
Agriculture in Africa is primarily rainfed, rendering the continent's agricultural sector very sensitive to climate change, fluctuating climatic conditions, and crop losses due to extreme weather events [1].The impact of climate change is already affecting agricultural productivity, particularly among impoverished smallholder farmers [2].Across the African continent, less than six percent of agricultural land is irrigated.For the next years, the demand for water will keep on rising to sustain productivity increases [3].The adoption of climate-smart agriculture might greatly enhance the agricultural sector by assisting African farmers in maximizing the potential pent up inside the small-scale agricultural system [4].Crop modeling supports agriculture in multiple capacities.They contribute to the comprehension of the interactions between the environment, crop, and soil, as well as pest management and natural resource management.Their contribution extends to evaluate the climate change impact on crop productivity, identify low-performing areas, and recommend agronomic practices to enhance profitability while promoting soil carbon storage [5,6].The core objective of this study is to offer an overview of the crop modeling technologies and their applications within the climate change context and the CSA framework in Africa, evaluate recent research advances in various crop growth simulation models, identify research gaps, and propose refined approaches to address climate change issues and sustain food security targets.The first part of the study is dedicated to the systemic analysis of crop modelling studies in Africa by examining the distribution of modelling approaches deployed and identifying crop models used to meet major agricultural challenges of climate change.The second part discusses the crop models applications followed by outlining research needs and proposing potential improvements.
Considering the present situation of the African agricultural sector and the crucial role of crop modeling tools in improving climate change mitigation and adaptation measures and promoting climate-smart agriculture targets, the systematic review answers the following questions: What is the recent status of crop modeling in the climate change context in Africa?The current gaps and research need in crop modeling, climate change adaptation, and mitigation in the climate-smart agriculture framework?And which opportunities and guidelines exist to address future research needs.Figure 1 illustrates general screening approaches and flow of identifying relevant literature which provides a detailed overview of the methodology used in this study.This technique was intended to evaluate systematic reviews that provide more efficient and less biased data.for scientists, stakeholders, and decision-makers.The articles used in this work were obtained through the Web of Science and the Scopus databases using the following keywords to refine the articles reviewed: "Crop modeling", "Climate change", "Adaptation", "Mitigation", "Climate-Smart agriculture", and "Africa".The review focuses on crop modeling studies on the African continent, the key words selected in figure 1 has helped to limit the scientific articles examined between 2000 and 2022 which is the period that includes most of the modeling studies.

Fig.1 Flowchart for the selection of literature
Following the initial results, 345 articles were identified from searching on Scopus, and 22 papers were obtained on the Web of Science.After deleting gray literature, presentations, keynotes, book chapters, non-English language papers, and extended abstract, the number of publications was limited to 248 articles left for further abstract and title scanning.In addition, some research articles were excluded due to low correspondence with the review objectives or the similarity of articles.Keeping the aim of this study, 140 relevant articles were selected and analyzed.A full list of related papers was obtained for further analysis.The synthesis process involved extracting and classifying pertinent data from selected papers to obtain conclusions.Data extraction technique includes determining and collecting pertinent data from the selected articles, such as (research objectives, major outcomes, research limits, model employed and required inputs, and calibration process).

Distribution of empirical and physical crop modelling studies in Africa:
Figure 2 describes the frequency distribution of physical and empirical crop models.Process-based crop models and empirical crop model studies accounted for 84% and 16% of these studies respectively, several crop modeling research in Africa employ process-based crop simulation models, such as DSSAT, APSIM, AquaCrop, and CROPSYST, due to their ability to simulate the effect of climate, soil, and crop management on crop development and predict potential yield.They are also well adapted to African agroclimatic conditions, as they can simulate a wide range of crops, soils, and climates.Analysis of variance, regression, and Pearson correlation models are widely employed due to their robustness and effectiveness in providing detailed information about variable relationships, contributing to reduced uncertainty.However, there is still a risk of introducing bias if important elements of the system are excluded by extrapolation of correlative relationships beyond the bounds of observed variability.The physical models rely on an explicit description of system behaviour based on the simulation of physical processes, where the empirical modelling approach involves correlative relationships in accordance with the mechanistic understanding, without a full description of system dynamics [7] The study revealed that maize is the most frequently used crop in climate change modelling studies in Africa.Despite its significant impact on food security and its adaptability to diverse soils and climates [8], maize has received less attention in climate change-related crop modeling studies in Morocco over the past decade compared to wheat, as indicated by the results.
Adaptation is essential to reduce the harmful impact of Climatic variability and is considered an urgent and necessary measure in the African context, to protect crop production and sustain the continent's food security, improve the productivity of mainly subsistence crops, and reduce yield gaps. African farmers are under increasing pressure to produce higher yields in the most vulnerable regions, the potential for implementing mitigation measures is rather low and adaptation is the major concern.On the other hand, reduced external input farming systems, which depend on the use of natural processes and organic fertilizer to ensure sufficient crop growth, reduce dependence on the supply of fertilizer and other inputs.In the long term, these systems are a valuable mitigation option that can enhance adaptation synchronously and that need to be developed locally [31].

Conclusion
Climate change and food security are among the major current development challenges.Agricultural production shows large annual fluctuations, mainly related to weather conditions and erratic rainfall.Moreover, the agricultural sector is missing many opportunities for sustainable development because climate change mitigation strategies have not been considered.72% of studies have contributed to adaptation strategies and reducing yield gaps in Africa, while only 8% of studies have been conducted on climate change mitigation and their trade-offs with adaptation using crop models under CSA.However, it is widely admitted that to attain food security goals through sustainable agricultural practices in Africa, implementing CSA models with a focus on the climate change mitigation component can greatly improve crop productivity, and farmers' incomes and help reduce greenhouse gas emissions.
Figure  3  indicates that crop modeling studies are focusing primarily on climate change adaptation strategies.The most used adaptation strategies are water and nutrient management, shifting planting dates, and incorporating new varieties, which are the important factors affecting crop yield.For example, Planting date shifting has been tested by several crop models in Africa as an adaptation practice to climate change for different crops including Sorghum

Fig. 3 .
Fig.3.Crop model applications to address climate change effects in Africa.