Thesis advisor: Noah P. Snyder / Centuries-long intensive land use change in the northeastern U.S. provides the opportunity to study the response timescale of geomorphic processes to anthropogenic perturbations. In this region, deforestation and the construction of dams following European settlement drastically altered the landscape, leading to the impoundment of sediment in mill ponds. This legacy sediment continues to be released into transport decades after a dam has been removed or breached. Geochemical tracers can help distinguish sediment sources and understand how sediment moves through a watershed. The South River in western MA is located in a formerly glaciated watershed, and these surficial deposits compose 98% of the area. It experienced two generations of damming, beginning with smaller mill dams in the 18th-19th centuries, followed by the construction of the Conway Electric Dam (CED), a 17 m tall hydroelectric dam in the early 20th century. Legacy sediment deposits from sediment stored behind mill dams cover 1.5% of the watershed area. The CED is located near the outlet of the river, providing a century-long depositional record for the watershed, during reforestation. I hypothesize that sediment mobilized from human activity will contain a different geochemical signature than glacial material, that recent erosion in the watershed is primarily from anthropogenic legacy deposits rather than from glacial age landforms, and channel widening is occurring in reaches of the channel composed of legacy sediment, rather than in glacially confined reaches. These hypotheses were tested through a two part investigation, consisting of a sediment tracing study using Hg, and a Geographic Information Systems (GIS) analysis of channel changes using aerial photographs from 1940 and 2014. Samples were collected from river bank exposures of 11 glacial deposits and four mill pond legacy sites. Two vibracores measuring 476 and 500 cm were collected in reservoir sediment stored behind the CED in 2013 and 2017, respectively. Hg concentrations range from 1-4 ppb in glacial sediment, 3-380 ppb in legacy sediment, and 2-18 ppb and 7-50 ppb in the two CED cores. I used Hg as a tracer to estimate percent contributions to the CED reservoir from each watershed source during the 20th century. Results from a sediment mixing model suggest glacial sources contributed 32 ± 15%, and legacy sediment deposits contributed 68 ± 15% during the 20th century. Based on 137Cs dates on the cores, high amounts of legacy sediment filled in behind the CED prior to 1953 (74 ± 35 %), and background erosion from glacial deposits dominated from 1953 until the reservoir was filled in the 1980s (63 ± 14%). GIS analyses using aerial photographs from 1940 and 2014 indicate that the channel did not significantly widen along any section of the river, however, increases in sinuosity (up to 12%) occurred in the legacy sediment dominated reaches of the channel, and minor increases (1-2%) occurred in the glacial reaches. Overall, these analyses show an increase in the amount of sediment released in the channel as a result of mill dams breaching through the mid-19th to early 20th centuries, and suggest a short recovery timescale response from this land-use change. / Thesis (MS) — Boston College, 2018. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Earth and Environmental Sciences.

Thesis advisor: Noah P. Snyder / Dam-influenced streambank morphology has not been studied extensively in rivers in deglaciated landscapes with high densities of colonial-era milldams. Fluvial restoration in the eastern U.S. often focuses on understanding pre-Colonial floodplain processes. Recent work by Walter and Merritts (2008a) in the Piedmont of the U.S. mid-Atlantic region suggests milldams significantly impact sedimentation by creating surfaces composed of post-dam legacy sediment that are often abandoned by the river and function as fill terraces. I analyze channel morphology and sedimentation patterns upstream of two breached dams on the Sheepscot River in mid-coastal Maine using lidar digital elevation models, historical aerial photographs, radiocarbon dating, and hydraulic modeling. In the past several decades, observable channel morphologic changes occurred at the two study sites: Maxcy's Mills dam (built in 1809, it was 2-m high and breached in the late 1950s), and at Head Tide dam (built in the 1760s, it is 4-m high and was partially breached in 1952). The Sheepscot River has a native population of Atlantic salmon, which is a federally listed endangered species. Understanding the existence and transport of legacy sediment has become an important component of habitat restoration efforts in the region. The goal of this investigation is to determine the extent and morphologic function of legacy sediment in order to better understand how historical dam sites affect channel morphology and sediment transport in a post-glacial, low-gradient river system. Field and remote sensing analyses indicate that surfaces (up to 2.65 m high) composed of mud and sand function as floodplains 1 km upstream from Head Tide dam and surfaces 90 cm high continue 2.5 km upstream from Maxcy's Mills. Analysis of seven radiocarbon dates from pieces of tree bark sampled from the stratigraphy (58-187 cm below the surface) of the two study sites suggest up to 1.8 m of sediment upstream of the two study sites was deposited within the past 300 years and is therefore a legacy of the dams. Quantification of the total volume of stored legacy sediment is on the same scale as volumes observed in the mid-Atlantic Piedmont region, leading to the conclusion that post-glacial rivers in northern New England store milldam sediment in similar fashion to streams analyzed in the Walter and Merritts (2008a) study. / Thesis (MS) — Boston College, 2013. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Earth and Environmental Sciences.

Thesis advisor: Noah P. Snyder / River channel morphology in northern New England depends on channel position relative to glacial geomorphology and history. This thesis considers three paraglacial Maine rivers: the West Branch of the Pleasant River (WBPR), a steep inland imposed-form tributary of the Piscataquis River, and the Narraguagus and Sheepscot rivers, two coastal low-gradient rivers. I use a simple model based on the Shields and Manning equations to predict median bed grain size in these recently deglaciated watersheds. The main objectives of this study are to: (1) understand how bedrock controls on the longitudinal profile and sediment inputs impact substrate grain size and channel morphology in the WBPR; (2) apply a model predicting substrate grain size based on digital elevation model (DEM)-derived geometric channel parameters; (3) compare the results from the high gradient WBPR to previously studied low-gradient coastal Maine rivers; and (4) explore the implications of my findings on channel and habitat restoration in paraglacial rivers. I use standard and lidar (light detection and ranging) digital elevation models (DEMs) and spatial analyses to measure channel parameters necessary to predict bed grain size and compare them to field measurements. Predicted bed grain size falls within a factor of two of the field-measured median in ~70% of the study sites. The model performs best in supply-limited alluvial single-thread channel segments with gravel-cobble lag deposit beds, and is less successful in transport-limited depositional segments with relatively fine beds and greater channel variability. Channel segments that are transitional between these two cases (intermediate channel complexity and grain size) are associated with intermediate grain size prediction accuracy. Model failures occur in segments that deviate from the single-thread gravel-bed channel type, and may indicate areas to focus restoration efforts. This study builds on previous research on low-gradient coastal rivers in Maine, and has wide application to future research or restoration projects concerned with sediment mobilization and fluvial ecology. / Thesis (MS) — Boston College, 2011. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Geology and Geophysics.

Submitted by Gizele Lima (gizele.lima@inpa.gov.br) on 2017-11-06T13:24:38Z No. of bitstreams: 2 20170911_tese_willian_v05_Final.pdf: 4868528 bytes, checksum: edb18b362ecc7e75c0532e2fd738afba (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-11-06T13:24:38Z (GMT). No. of bitstreams: 2 20170911_tese_willian_v05_Final.pdf: 4868528 bytes, checksum: edb18b362ecc7e75c0532e2fd738afba (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-07-14 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The world’s tropical forests, and the Amazonian Forest in particular, play an important role because they store between 193 ± 58 Pg and 228 ± 12 Pg of carbon and are facing intensive conversion to other land uses. There is a high level of uncertainty related to the quantification of this carbon reservoir and its emissions, in part due to the low density of field samples to characterize the natural variability. This research aimed to develop allometric equations for estimating total above and below-ground dry biomass for both trees and bamboo, apply these equations on forest inventory data, and test methods of extrapolation of the estimates to the landscape through remote sensing information. In order to adjust the allometric equations, we used the direct compartment method, roots (thin 2 mm < ∅ < 5 cm and thick ∅ ≥ 5 cm), trunks, branches (thin ∅ < 10 cm and thick ∅ ≥ 10 cm) and leaves of 190 trees with diameters varying between 5 and 92 cm; and 206 bamboo individuals (Guadua weberbaueri), subdivided in below- (roots) and above-ground (stems, branches and leaves) biomass. The basic wood density was determined in three trunk positions and in thick branches (∅ ≥ 10 cm) in 81 trees of different species with diameters varying between 11 and 90 cm. To determine forest biomass from remote sensing data, methods and density of LiDAR points were tested. The results showed that the allometrics patterns for estimating tree biomass in the Southwestern Amazon are different from other sampled regions in the Amazon. This fact may be related to lower tree height and wood density and higher water content in the fresh biomass. Were tested eight allometric models to estimate below-ground, above-ground and total biomass of individual trees in primary forest. Considering accuracy, practicality and costs, the use of the simple power equation involving only diameter (AGB tree = a × D b ) presented the best performance to estimate forest biomass. Bamboo biomass is an important component of the forest carbon cycle in a considerable part of the Southwestern Amazon. Were found a low allometric relation between bamboo dried biomass and its stalk diameter and height, a result distinct to those found by other authors, suggesting that there are different allometric patterns among the bamboo populations in this part of the Amazon. For the Amazon rainforest, an environment of low topographic variability, it is recommended the use of LiDAR point clouds with a density ≥ 2 m −2 to generate forest structure metrics and biomass estimation. To increase sample density is fundamental to improve the accuracy of forest biomass estimates. However, in order to contemplate spatial variability and access a large territorial extensions ecosystem such as the Amazonia, it is necessary to combine field data with remote sensing data as LiDAR. Open forests (+200,000 km 2 ) in the southwestern Amazonia are significantly different from forests in other regions of the Amazon. These differences can lead to disparities of up to 35 % in estimated forest biomass and consequently in carbon stocks and fluxes between forests and the atmosphere. To improve the accuracy of forest biomass estimates via LiDAR, consideration should be given to: (1) The quantity and size of the calibration plots; (2) the density of LiDAR points; and (3) the computation method. Keywords: forest biomass. allometric equations. LiDAR. Acre. / As florestas tropicais do mundo, em especial a Floresta Amazônica, têm um papel im- portante, pois armazenam entre 193 ± 58 Pg – 228 ± 12 Pg de carbono e estão sofrendo processos intensivos de conversão em outros usos. A incerteza associada à quantificação desse reservatório de carbono e suas emissões, é grande. Em parte, pela baixa densidade de amostras de campo para caracterizar a variabilidade natural. O objetivo deste trabalho foi desenvolver para a Amazônia Sul-Ocidental, equações alométricas para estimar biomassa seca total, abaixo e acima do solo em árvores e bambu, como também, aplicar a dados de inventário florestal e testar métodos de extrapolação das estimavas para a paisagem usando informações de sensoriamento remoto. Para ajustar as equações alométricas foi realizada estimativa de biomassa, pelo método direto dos compartimentos, raiz (fina 2 mm < ∅ < 5 cm e grossa ∅ ≥ 5 cm), tronco, galhos (fino ∅ < 10 cm e grosso ∅ ≥ 10 cm) e folhas de 190 árvores com diâmetro variando entre 5 e 92 cm; e de 206 indivíduos de bambu (Guadua weberbaueri), subdividindo em biomassa abaixo (raízes) e acima (colmos, galhos e folhas) do solo. Foi determinada, também, a densidade básica da madeira em três posições no tronco e galho grosso (∅ ≥ 10 cm) em 81 árvores de diferentes espécies com diâmetro variando entre 11 e 90 cm. Para estimar biomassa florestal a partir de dados de sensoriamento remoto foram testados métodos e densidade de pontos LiDAR. Os resultados mostram que os padrões das relações alométricas para estimar biomassa de árvores na Amazônia Sul-Ocidental são distintos de outras regiões amostradas na Amazônia. Este fato pode estar relacionado a menor altura e densidade da madeira e maior teor de água da biomassa fresca. Foram testados oito modelos alométricos para estimar biomassa abaixo do solo, acima do solo e total de árvores individuais em floresta primária. Considerando os parâmetros de qualidade, praticidade de uso e custo, a equação de simples entrada em potência, envolvendo somente diâmetro (AGB tree = a × D b ) teve melhor desempenho na estimativa da biomassa florestal, sendo preferível a utilização da mesma. A biomassa de bambu é um importante componente do ciclo do carbono florestal em uma porção conside- rável da Amazônia Sul-Ocidental. Foi encontrada baixa relação alométricas da biomassa seca de bambu com diâmetro e altura, resultados contrários aos encontrados por outros autores, sugerindo haver padrões alométricos diferentes entre as populações de bambu nesta porção da Amazônia. Em condições de floresta tropical amazônica em ambiente de baixa variabilidade topográfica recomenda-se o uso de nuvens de pontos LiDAR com densidade ≥ 2 m −2 para gerar métricas de estrutura da floresta e estimativa de biomassa. As floresta abertas (+200.000 km 2 ) na Amazônia Sul-Ocidental são significativamente diferentes de florestas em outras regiões da Amazônia. Essas diferenças podem provocar disparidades de até 35% na biomassa florestal estimada e consequentemente nos estoques e fluxos de carbono entre as florestas e a atmosfera. Para melhorar a acurácia das estimativas de biomassa florestal via LiDAR deve ser levado em consideração: (1) A quantidade e tamanho das parcelas de calibração; (2) a densidade de pontos LiDAR; e (3) o método de computação. Palavras-chave: Biomassa florestal. Equações alométricas. LiDAR. Acre.

Weather and climate forecasts are almost exclusively produced by computer models, which use atmospheric measurements as starting points. It is a well-known and joked-about fact that model predictions can be incorrect at times. One of the reasons this happens is due to gaps in our knowledge of atmospheric conditions in areas where measurements don’t often taken place, such as the mesosphere, which stretches from roughly 45-90 km altitude. A lidar is a device that can shoot out short bursts of laser light to measure things such as atmospheric thickness at a distance. From this information one can then derive the temperature in the upper atmosphere. Using temperature measurements taken by lidar at Utah State University (41.74° N, 111.81° W) and temperatures from three popular atmospheric models, a comparison is made covering the period 1993-2004 at 45 km altitude. This comparison demonstrates poor predictive capabilities of the models at the target altitude and suggests the need for integrating datasets such as lidar data into future models. The modeling community depends on real measurement comparisons to bolster the reliability and credibility of their own work, and the comparison done here is intended to highlight an area in need of improvement.

This thesis concerns the numerical simulation and the remote sensing of aircraft wake vortex flows. Due to its lift force, an aircraft releases large scale swirling flows (vortices) in its wake. As these vortices can impact significantly the trajectory of a following aircraft, their study is of great importance for practical applications concerning safety of air traffic management. The investigation carried here is twofold: it concerns, on one hand, the physics and the numerical simulation of aircraft wake vortices and, on the other hand, the technique to detect those vortices and measure their properties.The numerical simulation of aircraft wake vortices requires high order and energy conserving codes. Moreover, as aircraft wake vortex flows are turbulent, subgrid scale (SGS) models are required to perform Large Eddy Simulation (LES) of these flows. In the first part of this work, the numerical codes used (here spectral and high order finite differences) are presented and validated. Several SGS models are presented and their ability to perform LES of vortical flows, also in presence of a ground is assessed. In particular a new “multiscale” model with a natural wall damping behaviour has been developed and investigated: its performance is very good. In the second part, two flows relevant to the wake vortex problem are analyzed. The LES of a wake vortex system in a weakly turbulent atmosphere allowed highlighting the turbulent equilibrium state of such a flow. LES of wake vortices in ground effect, with and without wind, were also carried out. These simulations help to better understand the physics of wake vortices. In the last part, we simulate the remote sensing of aircraft wake vortices using a ground based LIDAR (Light Detection And Ranging) system. The aim of this LIDAR is to sense aircraft wake vortices and turbulent winds. As the LIDAR signals are simulated using realistic parameters and realistic turbulent flows, this work serves as support in the design of an airport based LIDAR system.

Although manipulation of the light regime is a common goal of silvicultural treatments, the specific light conditions created are poorly documented for many forest types and geographic locations. To help quantify effects of silivicultural treatments on light conditions, basal area, canopy structure, and photosynthetically active radiation (PAR), collected both instantaneously and across time, were measured in central hardwood forests following silvicultural treatments. These measurements were used to: 1.) investigate the magnitudes of differences in understory percent ambient PAR following implementation of shelterwood and thinning treatments; 2.) document the specific amount and variability of understory percent ambient PAR in shelterwood treatments (mean residual basal area=21 ft2/ac [4.8 m2/ha]), thinning (78 ft2/ac [17.9 m2/ha]), and untreated controls (18 ft2/ac[4.1 m2/ha); and 3.) Examine relationships between: basal area and canopy cover; basal area and measured percent ambient PAR; and canopy cover and measured percent ambient PAR. It was found that greater light levels resulted from greater canopy removals. Indexes of variability in light across time and among locations within a stand were higher in the shelterwood and thinning treatments than in the uncut control. Simple linear regression relationships were observed between basal area and PAR (r2= 0.8784 for instantaneous measurements, r2= 0.9697 for continuous measurements), and basal area and canopy cover (r2=0.8479). Such relationships provide a means for including light management in forest planning and application of silivicultural treatments.

The United States Forest Service (USFS) Forest Inventory and Analysis (FIA) program provides a diverse selection of data used to assess the status of the nation’s forested areas using sample locations dispersed throughout the country. Airborne, and more recently, terrestrial lidar (light detection and ranging) systems are capable of producing accurate measurements of individual tree dimensions and also possess the ability to characterize three-dimensional vertical forest structure. This study investigates the potential of airborne and terrestrial scanning lidar systems for modeling forest volume and aboveground biomass on FIA subplots in the Malheur National Forest, eastern Oregon. A methodology for the creation of five airborne lidar metric sets (four point cloud-based and one individual tree based) and four terrestrial lidar metric sets (three height-based and one distance-based) is presented. Metrics were compared to estimates of subplot aboveground biomass and gross volume derived from FIA data using national and regional allometric equations respectively. Simple linear regression models from the airborne lidar data accounted for 15 percent of the variability in subplot biomass and 14 percent of the variability in subplot volume, while multiple linear regression models increased these amounts to 29 percent and 25 percent, respectively. When subplot estimates of biophysical parameters were scaled to the plot-level and compared with plot-level lidar metrics, simple linear regression models were able to account for 60 percent of the variability in biomass and 71 percent of the variation in volume. Terrestrial lidar metrics produced moderate results with simple linear regression models accounting for 41 percent of the variability in biomass and 46 percent of the variability in volume, with multiple linear regression models accounting for 71 percent and 84 percent, respectively. Results show that: (1) larger plot sizes help to mitigate errors and produce better models; and (2) a combination of height-based and distance-based terrestrial lidar metrics has the potential to estimate biomass and volume on FIA subplots.

The design of a texture display for three-dimensional Light Detection and Ranging (LiDAR) point clouds is investigated. The objective is to present a low fidelity display that is simple to compute in real-time, which utilizes the pattern processing capabilities of a human operator to afford an understanding of the environment. The efficacy of the display is experimentally evaluated by in comparison with a baseline point cloud rendering. Subjects were shown data based on virtual hills, and were asked to plan the least-steep traversal, and identify the hill from a set of distractors. The major conclusions are: comprehension of LiDAR point clouds from the sensor origin is difficult without further processing of the data, a separated vantage point improves understanding of the data, and a simple computation to present local point cloud derivative data significantly improves the understanding of the environment, even when observed from the sensor origin.

The design of a texture display for three-dimensional Light Detection and Ranging (LiDAR) point clouds is investigated. The objective is to present a low fidelity display that is simple to compute in real-time, which utilizes the pattern processing capabilities of a human operator to afford an understanding of the environment. The efficacy of the display is experimentally evaluated by in comparison with a baseline point cloud rendering. Subjects were shown data based on virtual hills, and were asked to plan the least-steep traversal, and identify the hill from a set of distractors. The major conclusions are: comprehension of LiDAR point clouds from the sensor origin is difficult without further processing of the data, a separated vantage point improves understanding of the data, and a simple computation to present local point cloud derivative data significantly improves the understanding of the environment, even when observed from the sensor origin.