Agriculture that aims at good productivity and better land use can avail itself of technological systems to analyze the potential of soil. Called precision agriculture, this field is in constant evolution and stays in line with microelectronic progress through palmtops, software, global positioning sensors (GPS) and farming equipment, remote soil sensing and geostatistics. Some of these solutions have been brought together and have made the future of agriculture more attractive in studies conducted by professor José Alexandre Demattê, from the Luiz de Queiroz Agriculture College (ESALQ) at the University of São Paulo (USP). He and his group have developed a new type of digital soil map that facilitates the planning of planting and can be prepared faster than a conventional map. “We use satellite images, air photos and portable sensors simultaneously to obtain a more detailed soil map. This allows the farmer to choose the best land for planting, establish the legal forest reserves, rationalize the use of fertilizers and select the best varieties to increase the productivity of a given crop,” says Demattê.
“The soil maps currently prepared in Brazil have very few details, are expensive, and take time to be prepared because they demand hard work.” To prepare a more detailed and less expensive map, Demattê uses soil analysis obtained through reflectance, the energy reflected from the soil and captured in the form of electromagnetic radiation by the sensors located on land and on satellites, such as the United States’ Landsat and Aster satellites. “Electromagnetic radiation is associated with soil elements such as clay, sand, iron oxide, potassium, calcium, organic matter and minerals.” The soil has to be exposed for one to evaluate the surface layer of an area by satellite images. If the area is covered with vegetation, one must change the method, so that the shape of the terrain, such as soil elevation models, is used. This is generally obtained from shuttle radar topography mission (SRTM) system. This mission was conducted by the Endeavour space shuttle in 2000. “The objective in using satellite images is not to learn the soil classification, because the image only captures the surface layer, but, rather, to get more information that can be added to other data, allowing one to conclude what the likely soil type is.”
An optical sensor, which can be carried in a knapsack, is another tool used by the researcher in covered and uncovered areas. Not yet widely employed in agriculture, this device costs some US$60 thousand. “The optic fiber is pointed at the soil to capture the reflected energy. The data is later processed and mathematical models quantify and help draw a detailed map of the soil,” says Demattê. The sensor does not entirely substitute laboratory analyses conducted to find out the composition of the soil. “It allows more rational collecting of samples. For example, if it is necessary to collect 500 samples in a density of 1 per hectare (ha), the new method collects the same 500 samples, but only 150 samples would be sent to the lab and the rest would be quantified by the sensors (which reads the data in one second), which saves as much as 64% of the cost of soil analysis, as demonstrated in the master’s degree dissertation of Ramirez Lopez, who has a FAPESP grant and is a member of the group.”
Demattê proposes that all the techniques for drawing new maps be more closely integrated. “It is possible to add information from each type of equipment, such as the satellites, field sensors, elevation models, terrain shape, etc.” He adds that there are two scientific communities that rarely interact and this is reflected on the community in general. “One community consists of researchers in the field of remote sensing and the other comprises soil scientists. The members of one community do not use the soil analysis knowledge of the other community, whereas the members of the other community do not realize the real advantage of applying remote sensing.”
The integration of software for the new system is still not ready in order to help farmers. “This will be the next phase of our studies, when we will systematize the input of parameters to get an end result. We have already contacted other institutions to establish a work sequence and make the system available for farming in general.”
The research of professor Demattê has also moved on to environmental monitoring, using optical sensors to identify quickly cattle blood discarded illegally on pastures and near streams. Each bovine generates 15 to 20 liters of blood, which can be sold by slaughterhouses to companies that process it to make plasma and meal for animal feed. Demattê prepared a study to evaluate the changes that occur in the soil. “We noticed changes in the soil in the western part of São Paulo State.” To detect blood in soil, the researchers used samples with and without the product; the samples were analyzed by sensors in labs.
“We found that the soil undergoes many chemical changes, largely due to excess sodium, which modifies the contents of the nutrients and the pH.” The data indicates that bovine blood is discarded in unsuitable places and that the residues flow into rivers and streams, probably also contaminating the water tables. The professor was able to implement a remote sensing method to detect and evaluate the problem quickly. The group researching Soil Science Geotechnology at Esalq (CeoCis), coordinated by Demattê, is now working on scientific articles that will later become a report to be submitted to Cetesb, the São Paulo State Environmental Agency, that is responsible for overseeing the destination of animal sub-products.
1. Integration of multiple techniques for soil mapping (nº 07/55241-1); Type Regular Research Awards; Coordinator José Alexandre Demattê – USP; Investment R$ 100,381.20 (FAPESP)
2. Detection of slaughterhouse sub-products using remote spectral sensing in the ultraviolet, visible and infrared regions (nº 05/59691-6); Type Regular Research Awards; Coordinator José Alexandre Demattê – USP; Investment R$ 30,514.95 and US$ 35,050.00 (FAPESP)