This dissertation probes three issues of current interest in environmental and agricultural economics. The first paper provides an in-depth analysis of sedimentation management issue in large reservoirs. The paper provides a new model of sedimentation management and conditional on assumed primitives of the model, analyses different scenarios under which sedimentation removal may increase or decrease. The paper also provides insights on how temperature fluctuation, increased sedimentation arrival in the reservoir and change in the perception of large reservoirs among the public may affect the sustainability and management of the large reservoir. The second paper looks at the data from the Latin American countries to search for the presence of Environmental Kuznets Curve(EKC) in Latin America. The paper is also one of the earliest papers to use forestry data and semiparametric approach in finding EKC. The paper finds no evidence of EKC in Latin America as a whole, and in general finds that EKC is sensitive to the region of choice. The third paper carries out an an empirical investigation to test for the convergence of total factor productivity(TFP) of agricultural sector in the United States. The investigation does not find any evidence of convergence while looking at the U.S. state-level agricultural TFP at the aggregate level. However, it finds support for convergence within some of the clusters or within some of the regions. The paper takes a new approach in grouping states, which makes it different from other papers where ad hoc grouping of states was done. In this paper, such approach is abandoned in favor of a cluster analysis approach that relies on data to form "clusters". Cluster analysis approach finds that convergence in the regional level (cluster) does not improve significantly compared to the findings by a wellknown previously published study which didn't use cluster analysis approach .