No description available.

This thesis introduces the implementation of different supervised learning techniques for producing accurate estimates of soil moisture content using empirical information, including meteorological and remotely sensed data. The models thus developed can be extended to be used by the personal remote sensing systems developed in the Center for Self-Organizing Intelligent Systems (CSOIS). The dfferent models employed extend over a wide range of machine-learning techniques starting from basic linear regression models through models based on Bayesian framework. Also, ensembling methods such as bagging and boosting are implemented on all models for considerable improvements in accuracy. The main research objective is to understand, compare, and analyze the mathematical backgrounds underlying and results obtained from dfferent models and the respective improvisation techniques employed.

<p> Peatlands are an important component of the global carbon cycle, storing 23 g C m-2 yr-1 to comprise a global carbon pool of approximately 455 Pg. Peat drainage and harvesting results in removal of surface vegetation, thereby reducing gross photosynthesis to zero. Moreover, lowering the water table increases carbon oxidation. Consequently, peatland drainage and mining can reduce or eliminate the carbon sink function of the peatland. In the first part of this study, net ecosystem CO2 exchange was studied in a natural (NATURAL), two-year (YOUNG) and seven-year (OLD) post cutover peatland near Ste. Marguerite Marie, Quebec during the summer of 1998. Although the NATURAL site was a source of CO2 during the study season, CO2 emissions were 270 to 300% higher in the cutover sites (138, 363, and 399 g CO2-C m-2; NATURAL, YOUNG and OLD, respectively). Active restoration practices and natural re-vegetation of peatlands have the potential to return these ecosystems to net carbon sinks by increasing net ecosystem production (NEP) and therefore decreasing CO2 emissions to the atmosphere. Net ecosystem CO2 exchange in a natural (NATURAL) peatland and a partly restored peatland (REST) near Ste. Marguerite Marie, Quebec, was compared with a naturally re-vegetated peatland (RVEG) near Riviere-du-Loup, Quebec. Ecophysiological parameters indicate that the REST site was more than twice as productive as the natural LAWNS and three times as the RVEG site (GPmax=18.0, 8.3, and 6.5 g CO2 m-2 d-1, respectively). These results indicate that active restoration improves carbon sequestration over natural re-vegetation but that the net carbon sink function at both sites has not been restored. The presence of Sphagnum cover at the RVEG site resulted in a significant decrease in net ecosystem respiration (NER), indicating the potential for decreasing soil respiration at restored cutover sites through increasing the volumetric soil moisture content.</p> / Thesis / Master of Science (MSc)

Due to inadequate long-term and large-scale observation approach for observation of soil moisture across the globe, this study intends to unveil the importance of using simulated soil moisture fields from land surface models, forced with observed precipitation and near-surface meteorology in monitoring drought and formulating effective water management practices for continued production irrigation applications. This study shows that socio-economic and ecosystem effects can be determined by evaluating spatial-temporal changes in irrigation applications. Thus, it facilitates understanding of the importance of water management and how water, energy, and carbon flows protect our climate and environment. By using Soil Moisture Active Passive (SMAP) in monitoring soil moisture, the information obtained is critical in providing early drought warnings, particularly in those parts of the United States that experience flash agricultural droughts. Further, this study highlights that frequent and reliable soil moisture measurements from SMAP helped improve the predictive capability of weather and climate models.

Measuring soil moisture content (SMC) in mining waste rock is important for assessing and modelling hydrological processes which influence pollutant release. Here, an experimental setup containing mining waste rock is established to compare the performance of 4 Arduino capacitance moisture sensors to one single Time Domain Reflectometry (TDR) sensor. Furthermore, the performance of these sensors is evaluated in both sieved and unsieved mining waste rock. Fitted calibration curves are provided for both the TDR- and Arduino-sensors individually and in combination. These calibration curves are evaluated using the RMSE and R 2 of each curve and compared between sensors and soil texture. It is concluded that using more capacitance sensors significantly improves the fit statistics of the calibration curves and that using at least 4 capacitance sensors can enhance calibration curve fitting. For both the TDR and capacitance sensors, the calibration curves in sieved soil provided the best fit, meaning that soil specific calibration of sensors is recommended. On a sensor individual basis, the temporal precision of the TDR sensor was superior to each individual capacitance sensor. Use of 4 or more Arduino capacitance sensors may especially be justified in circumstances where the spatial variability of SMC is addressed by executing a large number of measurements. Here, the feasibility of the Arduino sensor system means that the use of these low-cost sensors, despite their reduced temporal precision, can be upscaled at relatively small costs.

Thesis (M.Sc. Agriculture. (Soil Science)) -- University of Limpopo, 2022 / The constant growth experienced by the coffee industry has led to the high-volume production of coffee waste worldwide. One of the main coffee wastes is spent coffee ground (SCG), a residue obtained after the ground coffee beans are treated under pressure. The present study was aimed to investigate the utilization of SCG to amend soil physicochemical properties. This study was conducted at Greenhouse Biotechnologies Research Centre of Excellence, University of Limpopo, South Africa, where the effect of various rates of SCG concentration in volume percentage (vol%) was tested for a period of nine months. The spent coffee ground residue was collected from four restaurants at Haenertsburg, and the application rates were 0, 5, 10, 20, 30, 50 vol%. To evaluate the change in soil physicochemical properties overtime, the incubation period was divided into four test periods namely T1 was after a month, T3 after 3 months, T6 after 6 months, and T9 after 9 months. Physicochemical properties including nitrogen (N), phosphorus (P), potassium (K), magnesium (Mg), calcium (Ca), total organic carbon (TOC), cadmium (Cd), copper (Cu), nickel (Ni), zinc (Zn), and lead (Pb), pH, electrical conductivity (EC), C:N ratio, large macroaggregates (LM), small macroaggregates (sM), microaggregates (m), unaggregated silt and clay (s+c), mean weight diameter (MWD) and soil moisture content (SMC) were quantified at the end of each test period. Results revealed that the interaction between incubation periods and various SCG application rates significantly (p<0.05) increased pHw, EC, MWD, LM, base cations and significantly decreased TOC, heavy metals, SMC, m, and sM. Spent coffee ground increased pHw and EC of the soil at all application rates and reached a maximum of 7.8 units at T6 in treatment SCG-5 and 202.30 S/cm at T9 in treatment SCG-50 above the control respectively. Total organic carbon increased by 548% above control in the highest treatment (SCG-50) at T1, but, however, started declining from T3 in all treatments across the incubation period. SCG’s highest application rates (SCG-20 to SCG-50) reduced the soil Cd toxicity (threshold of >2 mg/kg), but however, also reduced the availability of micronutrients (Cu and Zn) during the incubation period. At T9, Mg, Ca, K, and P increased from mean values of 55.9 to 77.9, 40.9 to 62.2, 77.4 to 112, and 22.0 to 30.0 mg/Kg above control in treatments with high application rates. LM increased whilst sM, and m decreased across the incubation period in all treatments. MWD increased by 46% at T1 and reached its maximum of 56% at T6 in treatment SCG-50 above control. Additionally, there was a positive relationship between LM and MWD. Soil moisture content however increased to 60.26% at T1 in treatment SCG-50 and decreased from T3 across the incubation period. Spent coffee ground has the potential to be used as a liming material, a chelating agent, and for water management in semi-arid areas. It retains and cycles nutrients and improves soil structure through aggregation. However, research should be done in field conditions to access the effectiveness of this residue. / NRF