THE USE OF EXTERNAL DATA BASES FOR THE ESTIMATION OF FUTURE PRODUCTION CONDITIONS
Although model predictions of future climate conditions are often bleak on the whole, it appears that in the agricultural sector changes can have either a positive or negative impact on production depending on the type of crop and the region and time period of activity.
For example, recent studies show that the climatic conditions for cotton crops can be substantially improved in the coming decades and already rank cotton among the "winners" of Climate Change [1]. The climate in Greece has a high probability of becoming drier, due to a combination of longer periods of high temperature and water scarcity. Warm tropical temperatures (above 30 degrees Celsius) are ideal for growing cotton crops as the deep roots of the plant make it resistant to drought and water scarcity [1]. However, despite the potential increase in irrigation costs that these climate changes may entail, it is estimated that higher cotton production can offset these economic losses and benefit farmers. It is estimated that future climatic conditions can lead to an increase in yield and an annual economic profit of about 87-103 million euros, increased by 29% compared to 2015 prices, provided that irrigation water will remain available at current levels and prices, which seems very unlikely [1].
Such climate information (temperature changes, precipitation, etc.), as well as their future projections, fall under the category of Big Data applications. In combination with Internet of things devices (sensors, drones, sensors) and Artificial Intelligence techniques, they can be used in the agriculture, livestock and fisheries sectors to predict future environmental conditions and assess risks or opportunities.
These two graphs show that for the Kilkis region there is an upward trend in temperature over the years. This will obviously have an impact either positively or negatively on future crops as climate conditions change. [Source: E-ON INTEGRATION S.A.]
In the agricultural sector, the use of the above can make a decisive difference to the way farmers work in the field and contribute to decisions to optimize production and reduce losses. Decisions on what to sow, when and where to sow, what management actions to take (fertilization, irrigation, etc.) and when to harvest are among other information that can now be taken with the help of intelligent Big Data applications that inform, prepare and alert farmers appropriately. At the same time, the use of climate models, appropriately "translated" and adapted for their own situation, can help them to identify risks or opportunities emerging due to Climate Change.
At E-On Integration we have already carried out a pilot project for a large company in the agricultural sector, collecting historical data and building production forecasting models by considering climate data provided by external databases. We took into account humidity, temperature (atmospheric and soil), precipitation and other data at a spatial scale of 10 km.
Comparison of performance values between
Neural Network Model (AI) and classical calculations
Use of neural networks to correct crop yield predictions taking into account climatic and other characteristics. [Source: E-ON INTEGRATION S.A.]
Using AI tools, we built a crop yield prediction model to monitor production trends by species, region and growing season to identify optimal growing conditions and maximize yields. Our goal here is to use the climatology and region-specific conditions to allow farmers to choose the appropriate sowing and harvesting seasons, the most suitable seeds for each region/seeding season, and to choose between quality or quantity. So by using models a wide range of options is offered in terms of company expectations to optimize their production and reduce losses due to adverse conditions or possible wrong choices.
[1] E. Georgopoulou, S. Mirasgedis, Y. Sarafidis, M. Vitaliotou, D.P. Lalas, I. Theloudis, K.-D. Giannoulaki, D. Dimopoulos, V. Zavras, “Climate change impacts and adaptation options for the Greek agriculture in 2021–2050: A monetary assessment”, Climate Risk Management, Volume 16, 2017, Pages 164-182, https://doi.org/10.1016/j.crm.2017.02.002.