Desenvolvimento de software para simulação do sequestro de carbono no solo

Abstract

Climate change is a global issue aggravated by the accumulation of carbon (C) in the atmosphere due to anthropogenic actions. Since the discovery of fossil fuel, its burning has been the main emitter of C to the atmosphere, contributing to raise of Earth's temperature. There are also other sources of C emissions with significant participation in climate change, such as land use change. For centuries, deforestation, burning and poor soil management in the agricultural system have led to a significant loss of forest biomass and soil organic matter, releasing C into the atmosphere in the form of carbon dioxide CO_2, methane CH_4, among other greenhouse gases. Computational models for soil carbon balance were created to evaluate the effect of conventional and conservationist management practices on soil C reservoir. However, most of these models were developed for climate and soil conditions in temperate regions, thus requiring adaptations to Brazilian soil and climate conditions. The objective of this work is to adapt the CQESTR model in a more robust and modern programming language, making its use available for the most current operational systems, Linux and Mac OS, besides Windows. This model was chosen because of its accuracy, simplicity and the need for few input parameters, besides allowing to simulate carbon stocks in the soil at a depth of 3 m. The updated model, in the Python language, was developed by applying object oriented programming techniques. In order to evaluate the new structure of the model, data were used from the systems: no-tillage and conventional tillage with disc plowing, heavy harrowing and scarification, on wheat and soybean crop rotation in the experimental field of Embrapa Soybean in Londrina/PR. The results of the Python model presented similarities to those of the original model, although inaccuracies were found in the second, such as the change in thermal time for simulations with double agricultural rotation per year and the lack of one day in the leap year. The update allows to simulate more than two crops per year, along with leap year implementation in decomposition processes. It also allows to read data from .xlsx and .xls files, making this procedure more agile and easy for the user compared to the original program. The web interface modernized the use of the model, allowing its use in any operational system. The interface in Python will facilitate future updates to the CQESTR model.