Welcome to the Geo-Resolution 2023 Student Poster Session. Students submitted posters on a broad array of research topics connected to geospatial sciences or using geospatial tools and technologies.
The award committee, led by ¶¶Ňőpro Professor Ness Sandoval, selected the following posters for the top three prizes:
- First place: Yi-Chieh Lee, Missouri University of Science and Technology, "Quantifying crustal deformation in Hispaniola using InSAR and Geospatial Analysis"
- Second place: Adam Becker, University of Missouri-St. Louis, "Crops and How They May Predict Conflicts"
- Third place: Ryan Kelly, ¶¶Ňőpro, "A Boosted Decision Tree to Predict Atmospheric River Vapor Transport with Teleconnection Patterns"
We would also like to give special recognition to the winners of the “audience favorite” voting:
- Jon Eman, University of Missouri-St. Louis, "Smart Garden approach for water conservation using an Automated Precipitation Aware irrigation (AuPAIr) system"
Congratulations to our 2023 Student Poster Session winners!
Below are all the posters submitted for the competition.
Poster 1: Crops and How They May Predict Conflicts
Authors: Adam Becker, Thomas Cholak, Joshua Meppiel, Chris O’Steen, Jacob Wilson
Abstract: “This research aims to weigh the effects of weather patterns in Nigeria over a period from 2017 to 2021 and analyze the effects on the overall stability of the region. We looked at ten of the main food crops in the region of Oyo, Nigeria and analyzed how the overall production of resources changed over this period. This data provides insight on future developments in agriculture and can be utilized to make future predictions on the stability of the overall country. Our team’s recommendation is to focus on crops which utilize less moisture and to utilize crop practices which conserve water in the region such as planting cover crops. In addition, we recommend for the region to promote the growing of native crops over foreign crops since they will be more adapted to local weather patterns and require less water; or planting strains of crops which are more drought-resistant. This will ensure that the land will have some protection against future drought, which is especially prevalent in the region as can be seen by Lake Chad losing 90% of its water since 1963 (Sow, 2017). Historical records indicate the dry season in this region has become longer and rain patterns have become less predictable during the wet season, with dramatic changes in soil moisture due to droughts and floods occurring in recent years (USAID, n.d.).”
Poster 2: Calculating Triangular Greenness Index on sUAS Mosaic Data
Author: Benjamin Krock
Abstract: “The Triangular Greenness Index can be calculated using standard RGB cameras, and so it can be applied to widely available drone data. We flew a mapping mission with a Mavic 2 Enterprise Advanced and created a mosaic using MetaShape. The Triangular Greenness index for the mosaic is calculated in Python and evaluated qualitatively.”
Poster 3: Modeling Mega-Flooding in California & the Central Valley Using GIS
Authors: Bonnie W. Dana, Dorris Scott
Abstract: “In the past year, California has experienced two major flood events. While these may seem like anomalies given the long-lasting drought, “geologic evidence shows that truly massive floods ... [mega-floods] have occurred in California every 100 to 200 years” (Ingram). The most recent mega-flood occurred in 1862 after years of drought. Heavy rain and snow in late fall through the middle of winter caused by multiple atmospheric rivers dropped many times the normal amount of rain in areas across California, turning the Central Valley into an inland sea. As the climate changes, mega-flooding may become more common, which could have major impacts on both lifestyle of Californians and on the national economy. Considering how much the United States depends on California’s industry (tech, real-estate, etc.) and its produce, primarily from the Central Valley, a repeat of the 1862 flood would be disastrous. This project considers a variety of factors that could lead to a major flood event to determine the level of flood risk across the state of California and separately in the Central Valley. Additionally, the workflow has been translated to a model so that the analysis can be repeated for alternative locations.”
Poster 4: Fueling Global Logistics: Geospatial Analysis of POL Ports
Authors: Christian Blue, Jessie Bleile
Abstract: “Ports are essential nodes in global supply chains, moving large volumes of products, raw materials, and commodities across the world. With much of the transportation sector reliant on Petroleum, Oil, and Lubricants (POL), ports with POL processing capabilities are important to ensure delivery of fuel for many industries, including Defense. This project will focus on the satellite imagery signatures of POL port features for different ports across the globe. Select ports will be analyzed with commercial satellite imagery to collect key features, such as POL/fuel piers, POL terminal boundaries, and new shoreline construction. Attributes collected will include: World Port Index number, sensor type, Image ID, Image Date, Berth Lengths, and New Construction. Esri’s ArcGIS Professional software will be used to collect features and attributes from Maxar satellite imagery. Multiple images of the same ports will be considered to account for different image parameters, such as sun angles, sensor look angles, and different times of the year. The geospatial features collected will be used to train object detection models (currently under development) for the National Geospatial-Intelligence Agency (NGA). Overall, this project will help NGA analysts quickly and efficiently collect information about POL and fuel at ports across the globe to aid planning and analysis of Global Defense Logistics.”
Poster 5: From the Gates of Kabul to the Gateway City: A Spatial Analysis of Afghan and Non-Citizen Immigrants in St. Louis, Missouri
Author: Elizabeth Salley
Abstract: “St. Louis, Missouri, is home to a foreign-born population of 6.9% (2021) and federally designated a preferred community for refugee resettlement. Most recently, after the Taliban takeover in August 2021, St. Louis, Missouri, welcomed over 700 Afghan evacuees. Refugees and non-citizens in particular confront significant challenges in their new U.S. host societies, including limited federal cash assistance and welfare benefits, expectation of immediate employment, language acquisition, health issues related to war and migration trauma, and job market access. This study utilized 2020-2016 American Community Survey data for the St. Louis Metropolitan Area to conduct spatial and bivariate spatial autocorrelations for the dependent variables of Afghan population; noncitizen population, and foreign born population; and independent variables of poverty; health insurance; education attainment index; Theil index; and median household income. Additionally, spatial regression was conducted to compare models for correlations between percent non-citizen and poverty rate. Results indicate local bivariate spatial autocorrelation for percent Afghan and education attainment index; local bivariate spatial autocorrelation for percent non-citizen and education attainment index, percent without health insurance, poverty rate, and median household income; and spatial error as the best regression model for describing the spatial relationship between non-citizens and poverty rate. Local and policy level recommendations are offered based on these results, as well as additional insights from a qualitative analysis conducted with 20 Afghan refugee evacuees in 2022-23.”
Poster 6: NeRF-based 3D Reconstruction and Orthographic Novel-View Synthesis Experiments Using City-Scale Aerial Images
Authors: Dylan Chapel, Edward Shang, Taci Kucukpinar, Joshua Fraser, Jaired Collins, Kannappan Palaniappan, Vasit Sagan
Abstract: “3D reconstructions of city-scale scenes can be used in remote sensing applications such as object tracking, bundle adjustment, shadow detection, and shadow removal. Point clouds are a standard 3D format used to represent reconstructions, with lidar being able to produce high-fidelity point clouds, but it has a limited ability to capture vertical details such as building walls. Current structure-from-motion techniques will capture information about all visible areas of a scene, but they require significant computing power and are prone to inaccuracies. This study primarily focuses on leveraging neural radiance fields (NeRFs) to address these challenges in 3D reconstructions of city-scale environments. NeRFs utilize neural networks to transform 2D images into a 3D representation, offering an innovative approach for more comprehensive scene captures. Our work aims to evaluate the performance of publicly available NeRF models in reconstructing complex cities and buildings, highlighting their capabilities and limitations.”
Poster 7: Smart Garden approach for water conservation using an Automated Precipitation Aware irrigation (AuPAIr) system
Authors: Jon Eman, Rekha Meyer, Rick Meyer, Sanjiv Bhatia
Abstract: “Water scarcity has long been a concern due to climate change and increased demand from growing populations. We designed the Automated Precipitation Aware Irrigation (AuPAIr) system using an IoT device that integrates sensors to detect environmental conditions such as soil moisture level, temperature, and humidity. It combines the sensed data with machine learning and geospatial predictive weather data to determine the time and frequency to water plants in an urban garden or greenhouse. We have used sunflower plants as indicators in our first stage of testing and data analysis. In AuPAIr, we have connected sensors to a Raspberry Pi, which runs a script to periodically sense data and acquire images of the plants, and then report that data to a cloud server. We will use OpenCV, PlantCV, and PyTorch along with the sensor and image data to train a machine learning model to determine the optimal point of irrigation before the plant is considered wilted. Once successful with sunflowers, we will create configuration profiles for other plants, soil types, and geographic regions.”
Poster 8: Methods in R-Language and Python for Geospatial Covariate Extractions Utilized in Disease Spatiotemporal Analyses
Authors: Joseph A. Emanuel, S. Chakraborty, A. Raghavan, S. Hesting, I. Jaster, R.K. Raghavan
Abstract: “The spatiotemporal distributions of wildlife diseases are closely linked with environmental and climatic factors. As such, a number of these “drivers” have explanatory potential, and in an epidemiological context they can be interpreted as “risk-factors” of these diseases. Chronic Wasting Disease is a prion disease affecting cervids (ex. white-tailed deer), present in over 30 states and provinces in North America. Various geospatial factors, such as land cover and soil type/condition have been suggested to impact the spread of the disease. Wildlife disease data are often imbalanced, and understanding the associations with these factors is complicated and requires a robust geospatial framework for rapid covariate data extraction, management and parallel-processing of statistical models. We developed a number of programming routines in R-language as well as Python to address this critical need. We demonstrate by utilizing the “sp”, “sf”, and “raster” “exactextractr” R-packages, among others, covariate extractions corresponding to reported disease locations can be either summarized by ad-hoc areal units and/or specific locations on a given geographic space. Further, these routines are readily coupled with machine learning (ML), statistical diagnostic and interpretative methods (e.g., Gaussian Process Boosting, Generalized Additive Model, and Log Gaussian Cox Method) for determining strength of covariate associations and patterns of spatiotemporal distributions.”
Poster 9: Geospatial Analysis of Chronic Wasting Disease in a Generalized Additive Model
Authors: Joseph E. Mosley, M. Yoo, A. Raghavan, S. Hesting, L. Jaster, C. Wikle, R.K. Raghavan
Abstract: “Chronic Wasting Disease (CWD) is a widespread infectious degenerative prion disease affecting captive and wild species of cervids (ex. white-tailed deer). It is present in over 30 states and provinces in North America. Various geospatial factors have been suggested to potentially impact the epidemiology of the disease; however, such heterogeneous geospatial factors, i.e., soil, land cover/land use, landscape metrics that may contribute a risk and further its control and management are currently unknown. From 2020-present, the One Health Research Lab at MU partnered with the Kansas Department of Wildlife and Parks to conduct disease surveillance for CWD and monitor the various drivers of spread in Kansas. We constructed Generalized Additive Models (GAM) along with environmental covariates derived from the USGS, and USDA-NRCS to evaluate the strength of associations with disease status. Additionally, spatial and temporal smooth functions were added to adjust for spatial/temporal autocorrelation in the GAM construct. Our analyses indicate that spatial locations (latitude and longitude) of animal locations is a significant driver for this disease (P-value: <2e-16), however, our current environmental covariates were not explanatory. This suggests that the disease probability rates are at least partially defined by the spatial heterogeneity and further refinement of environmental covariates (e.g., appropriate size and varying units of analysis, higher resolution geospatial data, etc.) are required. Additionally, analysis at a regional scale which share common geospatial features (e.g., watershed) may help identify geospatial covariate associations.”
Poster 10: Segmentation of Arabidopsis thaliana Rosettes and Leaves using Segment-Anything
Authors: Landon Swartz, Suxing Liu, Alexander Bucksch, David Mendoza-Cozatl, Kannappan Palaniappan
Abstract: “The SMART pipeline is a high-throughput phenotyping pipeline that processes images to observe leaf-specific traits within nutrient stress experiments in hydroponics. However, the pipeline's current K-means clustering segmentation method is slow and inaccurate for nutrient stress phenotypes. However, the foundational model in segmentation, Segment-Anything, offers a new method of high-resolution segmentation of complex geometries present in plant systems. A new method for segmentation of plant rosettes using Segment-Anything and Grounded Dino is present. Multiple methods for segmentation of individual leaves are present to observe the advantages and limitations of the prompt-based method of deep learning models. An analysis of the segmentation results from Segment-Anything shows a powerful method for providing statistically viable data for biological insight into novel plant traits under nutrient stress.”
Poster 11: Benthic Macroinvertebrates to Monitor Surface Drinking Water Quality: A National Security Issue
Author: Melissa Busby
Abstract: “Benthic macroinvertebrate results showing water quality in the lower 48 states of the United States 2008 – 2019. The spatial question I would like to answer is, “are the water systems healthier now than in the past 10 years?” and “where are there water quality issues in the United States, according to the amount of BMIs present?” Using knowledge from my water quality class and knowledge I’ve learned from literature online; I mapped out the number of benthic macroinvertebrates across the United States in major rivers and streams to determine where the water quality has either declined or gotten better over the span of data, I have collected from 2008-2019.
Benthic macroinvertebrates are the best biological assessment method for determining the water quality of a stream because they determine how healthy the stream is by the quantity in the water. They are sedentary and don’t migrate downstream. They also indicate short term environmental impacts because their life cycle is so short. Macroinvertebrates can rapidly and easily examine the water quality on the spot. They have a range of tolerances of pollution and trophic level, so if certain macroinvertebrates are present, there is an expectation that a certain pollutant is present. They are easy to sample.
Scientists go out into the field, use a HESS sampler, and collect samples of macroinvertebrates from the stream. Then bring back the samples to the lab and study the different metrics to determine quantity and type.
According to the USGS, “The taxonomic composition and relative abundance of different taxa that make up the benthic macroinvertebrate assemblage present in a stream have been used extensively in North America, Europe, and Australia to assess how human activities affect ecological condition (Barbour et al. 1995, 1999; Karr and Chu 1999)”. BMIs (Benthic macroinvertebrates) can help distinguish a healthy river from an unhealthy one. The quick collection of “taxa” and observations of the amount and what type determine if the river/stream is healthy, and if the soil erosion rate is heavy or not. Typically, the more taxa, or BMIs the stream or river has, the healthier it is, but there are certain types of BMIs that can make one conclude that there is a certain pollutant present. I collected data from various resources to show a broad view of the lower 48 states and the amount of BMIs in rivers and streams 2008 versus 2019.
According to the Environmental Protection Agency’s Natural Aquatic Resource Survey, “The purpose of these indicators (count of BMIs) is to present the complex community taxonomic data represented within an assemblage in a way that is understandable and informative to resource managers and the public.” Using the EPA’s data collected from 2008 to 2019 I was able to visually show informative and understandable information spatially, that is currently only seen via an Excel spreadsheet.”
Poster 12: Early detection of underground natural gas leak from hyperspectral imaging by vegetation indicators and deep neural networks (DNN)
Authors: Pengfei Ma, Tarutal Ghosh Mondal, Zhenhua Shi, Genda Chen
Abstract: “Early detection of underground natural gas release in pipeline transmission can not only reduce the greenhouse effect but also prevent natural hazards. The timely identification of natural gas emissions remains difficult due to the limitations of current techniques. Vegetation indicators (VIs) based on hyperspectral imaging from remote sensing have been defined to monitor health deterioration to represent the stress occurrence due to natural gas leaks, but the sensitivity and separability of the VIs remain questionable. A field test was arranged to use ground vegetation to sense the underground methane emission. Hyperspectral reflectance of vegetation was obtained weekly at the plant scale and the leaf scale over two months of treatment for stress detection with VIs and DNN. Results indicated that plant pigment-related VIs, MCARI, has the best sensitivity but low separability to distinguish methane-stressed grasses. For specialized natural gas VIs, OSAVI and VMDI demonstrated higher sensitivity and separability in early detection of methane leaks, which discriminated vegetation stress after 21-day methane injection. DNN yielded an acceptable stress presence at an accuracy of 98.4% after a 3-week methane treatment. And discriminative analysis indicates that spectra increase in VIS and decrease in NIR due to methane exposure.”
Poster 13: Using Image and Spectral Sensors in an IoT Edge Sensor Approach to Reduce Contamination in Recycling
Authors: Pranit Vaikuntam, Orhun Aydin
Abstract: “An increasingly abundant consumerist lifestyle coupled with an increasing population poses challenges for the current recycling and waste management infrastructure. At present waste management operations are performed without any data that allows operational foresight. This poster showcases the feasibility of use of edge sensing to identify recyclability of solid waste. By determining which bins are contaminated and then avoiding those bins reduces the operational uncertainty of recycling operators. In this project, an edge sensor approach is proposed with spectral and image sensors sensing the materials entering a recycling bin. Additionally, an IoT system is proposed using AWS that stores data in the cloud. Preliminary results from the sensor data are showcased in their ability to determine the recyclability of solid waste. Sensors are tested in their ability to distinguish objects with Linear Discriminant Analysis, results show that the sensors have potential. This research is performed at the AI-CHESS Lab at SLU led by Dr. Orhun Aydin, in collaboration with the Rockwood School District.”
Poster 14: Intercropping study to facilitate GIS monitoring of agriculture and increased food production
Authors: Randi Jacobs-James, Tannea Brown, Jayashree Balakrishna, Odesa Weatherford-Jacobs, Rekha Meyer
Abstract: “Food insecurity is projected to increase in the face of climate change, and this is especially true for vulnerable populations. Food productivity is projected to be impacted by climate change which will affect urban and rural areas. Reducing food insecurity requires practices that increase crop yield, be adaptable to weather extremes and reduce carbon footprint of large agricultural tracts. Intercropping, the practice of growing two different crops together has advantages in terms of growing shade tolerant and intolerant plants and combining this design with “nitrogen fixing” and “non-nitrogen fixing” plants. This can reduce the need for fertilizer and reduce greenhouse gases resulting from decreased production of fertilizer. Our research question asks if there are areas in our state and city that are designated as food deserts and explores techniques to mitigate impacts of climate change on food production. Our experimental hypothesis asks whether (a) shading significantly lowers the chlorophyll content to deter intercropping and (b) whether chlorophyll measuring instruments like Dualex and Spad 502 give consistent measurements of chlorophyll and nitrogen levels. Experiments included a short crop grown under the shade of taller bean (nitrogen fixer) and tomato plants (non-nitrogen fixer). Chlorophyll levels using a Dualex instrument were measured in leaves of intercropped versus control plants. Results show that reduced chlorophyll levels in shaded plants may be offset by increased nitrogen levels. Ground truth data from this research can be set up for GIS analysis to allow for comparisons to other intercropping sites and for comparisons with drone data. ”
Poster 15: A Boosted Decision Tree to Predict Atmospheric River Vapor Transport with Teleconnection Patterns
Authors: Ryan Kelly, Orhun Aydin
Abstract: “Atmospheric Rivers (ARs) sustain the water resources of the Western United States, are prone to substantial interannual variability, and are modulated by intraseasonal climate variability. The extent to which modes of climate variability modulate storm characteristics is an active area of research with potential to extend operational meteorology from days into weeks. We apply an automated detection algorithm to the North American Regional Reanalysis to build a climatology of cool-season, maritime-originating, West Coast atmospheric rivers spanning 1979-2013. Using a gradient-boosted regression tree model, we combine the AR climatology with a suite of teleconnection patterns, ENSO, MJO, QBO, PNA, and PMM, to evaluate the relationships between an AR’s vapor transport and its surrounding climate modes. We also evaluate the varying importance of certain teleconnection patterns by grouping AR tracks with similar geography. We find that an atmospheric river’s water vapor transport is modulated by a combination of its geography and the underlying large-scale climate background state. In particular, the source region of an AR storm track, its distance traveled, ENSO status, and the strength of the Pacific-North-American Pattern determine the strength of an AR’s water vapor transport. This research is performed at the AI-CHESS Lab at SLU led by Dr. Orhun Aydin.”
Poster 16: A tale of two neighborhoods in the city of St. Louis: Health disparities in urban communities with food deserts
Authors: Shamari Long, Adonni Scott, Odesa Weatherford-Jacobs, Rekha Meyer
Abstract: “Food deserts are defined as areas where residents have low access to stores selling healthy and affordable foods. Urban areas with food deserts are markers for more than socioeconomic injustice, they impact health outcomes for residents living within them and often intersect with race. Our objective is to use ArcGIS to explore the interrelationships between health and food access in communities designated as food deserts. We focused on two neighborhoods, Fairground and Lafayette Square, approximately four miles apart in the city of St. Louis and looked at rates of asthma, COPD, diabetes, heart disease and cancer among residents of both neighborhoods. Data from Missouri Environmental Public Health Tracking system for 2103-2015, showed that incidence of chronic disease, as measured by visits to ER, was higher in the 63107 zip code (Fairground) compared to 63104 (Lafayette Square). Incidence of asthma, COPD, diabetes, heart disease, were 1.8, 1.9, 3.4 and 1.7 fold higher respectively in Fairground. Trend line analysis for 2006-2008 versus 2013-2015 showed that rates of all five chronic diseases increased in Fairground (63107) but decreased in Lafayette Square (63104). Deaths due to diabetes and cancer increased in Fairground but decreased in Lafayette square for the two time intervals. Our results expose health disparities among residents living in food deserts. This type of study using ArcGIS could be used to empower community sustaining entities, to establish in the area and provide a healthy social environment for its residents.”
[https://docs.google.com/presentation/d/1NuaUZ21tdhG_PEOD9IWepKxTWZ7IyBfS/edit?usp=sharing&ouid=106416878517976538988&rtpof=true&sd=true] Poster 16 Link
Poster 17: Reimagining COVID-19 Map with CARTOGRAMS
Author: Supria Sarkar
Abstract: “The COVID-19 pandemic has emerged as one of the most significant crises of the modern era, affecting millions of people worldwide. United States affected badly by the pandemic which possess a long-lasting impact on the economy and society. Visual representations of this data help to gain a comprehensive understanding of the pandemic's spread and severity across the country. A cartogram is a thematic map of a set of features, in which their geographic size is altered dramatically to be proportional to a selected variable in order to highlight another variable. While COVID-19 maps at state level can say a lot about its spatial distribution, the focus can be lost due to the biasness of massive areas for some states. Therefore, three popular state-level cartograms will be investigated to reimagine the COVID maps at state-level for the United States. This information can also be utilized to inform public health policies and interventions, with the potential to mitigate the pandemic's impact in the future. Overall, this project can contribute to the public health response to the COVID-19 pandemic and provide valuable insights for future infectious disease outbreaks.”
Poster 18: A Down to Earth Approach For Camera-To-World Map Georeferencing Using SFM
Authors: Timothy Krock, Jaired Collins, Joshua Fraser, Hadi AliAkbarpour, Kannappan Palaniappan
Abstract: “We explore a method to georegister dense 3D drone models as well as the images that were used in the model. This can be due to noisy, biased, or nonexistent metadata in images. We do this by aligning the raw imagery with a satellite image with known coordinates, such as Google Earth.”
Poster 19: The Dark Side of Light: Geospatial Analysis and Anatomical Investigation of the Effects of Light Pollution on the Maternal Paraventricular Nucleus During Pregnancy
Authors: Inaya Smith, Trena Harris, Carmel Martin-Fairey, Ph.D.
Abstract: “Exposure to light pollution during pregnancy has been associated with poor gestational, maternal, and fetal health outcomes in mammals. The hypothalamic paraventricular nucleus (PVN) plays a critical role in regulating neuroendocrine changes during pregnancy in rodents. Melatonin, a hormone involved in circadian rhythm regulation, is produced by the pineal gland in response to darkness and has been shown to play an important role in fetal development and maternal health during pregnancy. In this study, we used Geographic Information System (GIS) analyses to identify areas with high levels of light pollution and map them against data on adverse pregnancy outcomes, such as low birth weight and preterm birth. We found that there is a clear relationship between exposure to light pollution and poor pregnancy outcomes. We then investigated the effects of a light pulse during rodent pregnancy on clock gene expression and c-fos activation in the PVN. Previous findings suggest that exposure to a light pulse during pregnancy can affect clock gene expression in the PVN of rodents, potentially through alterations in melatonin production. These results will provide important insights into the potential effects of light pollution on maternal and fetal health outcomes, highlighting the need for further research in this area. Moreover, the study further supports the critical role of the PVN in the regulation of pregnancy-related neuroendocrine changes, and the influence of melatonin production on the PVN function during pregnancy. Understanding the molecular mechanisms underlying these effects may have important implications for mammalian maternal and fetal health.”
Poster 20: Quantifying crustal deformation in Hispaniola using InSAR and Geospatial Analysis
Authors: Yi-Chieh Lee, Jeremy Maurer, Rishabh Dutta
Abstract: “Earthquakes are one of the most devastating natural disasters, causing significant damage and loss of life. Understanding when and where they will occur is crucial to mitigate their impact. Therefore, the study of crustal deformation is essential, as it unveils the intricate mechanics of our planet's dynamic processes, offering critical insights into seismic activity, landscape evolution, and hazard assessment. In this research, we used Hispaniola as the research area, located between the North American and Caribbean Plates. We use geodetic measurements, including GPS and InSAR, to estimate the surface velocity field of Hispaniola with horizontal and vertical components, which provides a more comprehensive understanding of the strain accumulation and deformation patterns across the island. Meanwhile, we could infer the subsurface geometry, especially the geometry of the subduction zone, and fault slip rates by analyzing the surface velocity field. Preliminary results indicate that the maximum shear strain rate is concentrated on the northern and south-western sides of the island, coinciding with two major strike-slip faults, the Septentrional-Oriente fault zone (SOFZ) and the Enriquillo-Plantain Garden fault zone (EPGFZ). The dilatation rate result varies in different methods, but the trend indicates that the shortening rate dominates most of the region. Incorporating geodesy measurements enhances our grasp of strain accumulation, aiding slip rate inference and deformation modeling. This insight aids in understanding subduction transitions in the larger scale plate boundary region.”