Chapter 7 - Automated real-time forecasting of agriculture using chlorophyll content and its impact on climate change

Abstract

The core theme of this project is to assess the economic impact of climate change on Indian agriculture. Climate change is caused due to the emission of greenhouse gases like carbon dioxide (CO2), methane (CH4), and nitrous oxide from various industrial sources. Neyveli, being the source of heavy megawatt-generating stations, let out flue gases, which contain CO2, carbon monoxide, oxides of sulfur, CH4, and oxides of nitrogen. These harmful gases are responsible for depletion of the ozone layer, which has a significant effect on variation in weather and agricultural output and sometimes even produces acid rainfall. Considering the probable effects of climatic change on agriculture has motivated a vital change in the yield of agricultural products, livestock yields and also changes in the food production pattern and prices. This estimation of chlorophyll content can be done by extracting green colored pixels from the satellite images or images captured by the vision sensors and soil moisture sensor placed in the Indian agricultural area. These images are preprocessed for noise removal using edge detection technique. From the preprocessed images, feature descriptors like histogram of oriented gradients (HoG) are extracted. The HoG values are fused with the information gathered from soil moisture sensor. The extracted features are reduced using principal component analysis (PCA). The feature set is thereafter used as inputs to artificial neural networks using feed-forward structure trained with backpropagation algorithm (BPA). These estimates done using data analytics will lend a helping hand to the farmers to adapt themselves to the year within annual weather shocks. It can be inferred that the estimates, derived from short term, are capable of predicting the short- and medium-term impacts of climate change, which would direct the farmers to adapt rapidly to the changing climatic conditions. These short- and medium-term impacts of climate change are found to reduce the agricultural productivity by 4%–6% and 6%–9%, respectively. Hence it is inferred that the climate change entails significant impact on the revenue of the Indian economy until and unless the farmers can promptly identify and adjust to decreasing rainfalls and increasing atmospheric temperatures. The first challenge lies in analyzing the satellite images of the farmlands using efficient image processing algorithms to extract useful and meaningful information. This data extracted would be of a very large quantity and needs to be handled using some data analytics algorithm like BPA, whose prediction efficiency will be determined and also validated. The second challenge lies in mapping the emission of greenhouse gases with the images of the farmlands under three categories, namely, highly productive farmlands, medium productive farmlands, and less productive farmlands and correlating the yield of farmlands with respect to emission levels of greenhouse gases in particular environment under study.