Effects of meteorological data on the development of net blotch of spring barley in Turkey


ARSLAN R. S., Akci N., Celik D., DAŞBAŞI B., Ozaydin K. A., DAŞBAŞI T.

9th International Symposium on Innovative Approaches in Smart Technologies, ISAS 2025, Gaziantep, Turkey, 27 - 28 June 2025, (Full Text) identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/isas66241.2025.11101898
  • City: Gaziantep
  • Country: Turkey
  • Keywords: Artificial Neural Network (ANN), effect analysis, meteorological variables, Net Blotch, Spring barley
  • Kayseri University Affiliated: Yes

Abstract

Ensuring sustainability and productivity in plant production is of paramount importance, and plant diseases represent a significant challenge in this regard. Meteorological changes due to climate change cause the dynamics of plant pathogens to change and plant diseases to increase. Barley, which is important as the main grain in the production of animal feed, beer and other beverages, ranks fourth in the world after wheat, rice and corn in terms of its share in cultivation. Net blotch disease is one of the most devastating leaf diseases in the world and causes major decreases in grain yield and quality. Within the scope of this study, the effect of 9 different meteorological data collected from Turkey for the years 20212024 on the formation of the disease was analyzed comprehensively using the ANN network. Normalized data with Min-Max were used in the training and testing processes of an ANN network that can detect the disease with high success. An ANN network with metric measurements resulting in high performance and the relevant non-linear activation function were obtained. Then, the effect of 9 different inputs on the disease value was investigated with the help of the activation function. For this purpose, two different random days corresponding to the summer and winter months when there is no disease in the dataset were taken as basis. With the activation function, first the disease value for each day and then the disease value when each input is maximum, respectively, was calculated. Thus, the effect of the change in the input on the change in the disease value when each input is maximum compared to the case where there is no disease was calculated as a percentage. The ensuing computations yielded the following findings: (1) the climatic parameters of the relevant region, which are related to the temperature, minimum, maximum, and average temperature, sun exposure time, dew point temperature, have a positive effect on the formation of the net blotch disease, (2) the humidity and actual vapor pressure parameters have a negative effect, and (3) the rainfall and wind speed parameters have no significant effect since the effects obtained in the summer and winter periods do not confirm each other.