SCIENCE@home

Enhancing Climate Resilience through Sustainable Agricultural Intensification: An Attention-Based Neural Network for Climate-Smart Agriculture Modeling in Ghana.

T. Gadafi,Seini Toufic,Albert Awortwi Sagoe,Liang Decui,Hira Anwar

Published 2025 in Proceedings of the 2025 International Conference on Smart Agriculture and Artificial Intelligence

ABSTRACT

Ghana's agricultural sector, which is important for the country's food security and livelihoods, is becoming more and more vulnerable to climate change effects like changing temperatures and irregular rainfall patterns. This study looks into whether combining Sustainable Agricultural Intensification (SAI) with Climate-Smart Agriculture (CSA) methods could make farming more resilient to climate change. We developed a neural network model with an attention mechanism that could predict crop yields using localized climate data from 2020 to 2024, based on temperature and precipitation inputs. The study used a parametric yield model and created a Climate Resilience Index (CRI) to measure how stable things are when the weather is bad. The experiment showed that agricultural yields dropped dramatically without Climate-Smart Agriculture (CSA), especially in years with bad weather. On the other hand, yields with CSA stayed pretty stable. The CRI values ranged from 0.95 to 1.15, which means that CSA techniques generally made crops more resistant to damage. The neural network training quickly reached a point of convergence, and the predicted yields were very close to the actual yields, which confirmed the model's accuracy. The correlation study confirmed that temperature has a negative effect on yield and precipitation has a positive effect on yield. Visual patterns and distributions also show that climatic conditions and yield responses can change. The study shows that Climate Smart Agriculture (CSA) can help make farms more resilient when used with modern modeling techniques. To protect future food systems, Ghana needs to improve its Climate-Smart Agriculture programs, improve its climate data systems, strengthen its agricultural extension services, and add resilience modeling to its national planning frameworks.

PUBLICATION RECORD

  • Publication year

    2025

  • Venue

    Proceedings of the 2025 International Conference on Smart Agriculture and Artificial Intelligence

  • Publication date

    2025-06-13

  • Fields of study

    Not labeled

  • Identifiers
    DOI 10.1145/3767624.3767648
  • External record

    Open on Semantic Scholar

  • Source metadata

    Semantic Scholar

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-25 of 25 references · Page 1 of 1

CITED BY