​Modern sugarcane agricultural management tools

With computational evolution over the years, sugarcane agricultural management has now gained strong allies with the use of modern techniques such as mathematical modeling associated with the use of databases and satellite images. By combining these techniques, a powerful tool is created capable of assisting in decision making, always aiming to obtain a correct estimate of productivity based on the climatic conditions of each year. This is of great value for crop planning, as productivity predictions are normally made based on observations and experience of technicians through their “mental model”. The “mental model” is understood as all the experience and observation acquired over time by the technician. It can be seen that training a technician to carry out productivity estimates requires time, unlike a mathematical model which, depending on the database with reliable productivity and climate information, can be developed quickly. An additional difficulty in the estimate made by the technician is associating the climatic extremes that occurred throughout the development of the sugarcane field and evaluating the impact on final productivity. In this case, mathematical modeling would return productivity values ​​based on climate scenarios that occurred with greater precision when compared to the “mental model”.

PREVCLIMACANA Project

The PREVCLIMACANA project began in 2007 within the IAC with the aim of evolving and developing the PREVCANA model to estimate sugarcane productivity, and the current stage of development after nine years of calibration and use in the units comes advancing precisely with the units' database and the current possibility of using satellite images.

The PREVCANA model simulates the potential growth of sugarcane based on plant photosynthetic and soil-climatic parameters of each production site, including solar radiation, leaf area index, extinction coefficient, photosynthetic rate, temperature, respiration, age of plant, water availability and partition of photoassimilates. With the history of each productive area extracted from the database, the growth potential based on the climatic conditions of each year is simulated, enabling the construction of normal, favorable or unfavorable productivity scenarios in the harvest. This way, we have an estimate throughout the harvest for each productive area.

Applied example

Using the PREVCANA model and the database of a plant located in the region of Ribeirão Preto/SP - 2015 harvest, the growth potential based on the climatic conditions of the agricultural year is simulated, enabling the construction of normal productivity scenarios (Figure 1), favorable or unfavorable in the harvest.

Figure 1 presents the estimate throughout the 2015 harvest for each productive area of ​​this unit in the Ribeirão Preto/SP region. The spatialization of this estimate using the model can be done by joining satellite images using the vegetation index as an example.

Figure 1. Estimated productivity (TCH – tons of sugarcane per hectare) generated by the PREVCANA model using the database. The vegetation index can be used to vary productivity within the harvest block.

In Figure 1, we see areas with shades of darker green indicating estimated yields in the range of 150 to 100 tons for areas of sugarcane plant and second cut and areas with shades of yellow and yields in the range of 99 to 70 tonnes for third and second cut cane. fourth cut, and areas with shades of orange indicating productivity below 70 tons for fifth cane and other cuts where these are areas that require special attention, whether due to problems in the stand or even related to management, which depending on the distance from this productive area to the processing unit milling, is not economically sustainable and measures to reform or replant these areas must be taken. Therefore, using mathematical modeling, database and image techniques together, excellent results are obtained in crop estimation with errors below 3% in validations between estimated and realized yields (Figure 2).

Figure 2. Validation between estimated versus realized productivity.

The combination of mathematical modeling with the database generates considerable advances in estimating sugarcane productivity with reliable results in validating estimated and realized productivity.

The next steps in the evolution of productivity estimation are focused on maturation estimation models associated with satellite imagery. A reinterpretation of maturation estimation models in association with satellite images is already being tested and represents a new way of associating maturation modeling with satellite images and the variations that exist within production environments.