GIScience and Remote Sensing



  1. Why GISciences/Remote Sensing?

The growing complexity of environmental problems is creating a need for scientists with rigorous training and interdisciplinary background in natural science and social science to protect our agro-ecosystems and human health. It is now globally recognized that climate change, global warming, and impacts on biodiversity and marine life are not just national problems, but global problems as well. By addressing diverse environmental issues today, GIScience and remote sensing become the essential science that helps me execute these research practices. Using GIScience and remote sensing can help us discover and deal with environmental issues. With GIScience and remote sensing, we can respond rapidly when flash disasters take place, for example, flash drought. For chronic environmental issues, such as drought, GIScience can provide a long-time series analysis mechanism, which enables us to track the dynamic changes over years or even decades. This is quite important because it provides us a basis for problem analysis and decision-making.


(1) Geospatial Modeling

Population Spatial Distribution

Soil Moisture Semi-variogram

(2) Time-series analysis

Soil Moisture Time Series

Coral Reef Bleaching Index Time-Series Prediction

(3) AI/ML, Big Data Analytics and Deep Learning

Random Forest for Soil Moisture Modeling

Neural Network Model for Crop Yield

2. Modern Remote Sensing: UAV Remote Sensing and Proximal Sensing

While the traditional investigation-based approach reveals spatial patterns from a regional scale, remote sensing, and spatial data science help address environmental issues from modeling/monitoring approach from even macro scales. Modern remote sensing technology provides us images with higher spatial resolution, higher spectral resolution, rapid earth surface coverage, and lower cost, and therefore enables modern geographers to look at the phenomenon from both local and large scales, which grants us the power of spatial thinking to discover big pictures and answer big questions. With the technologies including novel remote/proximal sensing (satellites, CubeSats, Unmanned Aerial System (UAS), LiDAR, robotic automated imaging system, etc.), spatial analytics, big data analytics, High-performance computing (HPC), and cloud computing, I will practice my skills to examine geographical phenomenon at multi scales (macro/intermediate/micro) using a multisource data science and multi-level modeling (or, hierarchical modeling) approach. My research framework will be based on large-scale environmental modeling (using satellite), extended to intermediate-scale ecosystem health monitoring (using UAV), and connected with leaf-scale plant phenotyping (using proximal sensing).

(1) UAV Remote Sensing for Precision Agriculture (crop disease monitoring, leaf nitrogen modeling, plant stress detection, etc.

(a) Resonon Pika XC2 Hyperspectral Camera

(b) ArduPilot Mission Planner for Autonomous Flight Control

(c) LiDAR Unit on a DJI Matrice 600 Pro

(d) DJI Matrice 600 Pro Hexacopter

(e) LiDAR Mapping Showing the Structure and Surface of Switchgrass Canopy

(f) A Map of NDVI Generated from Multispectral Data

(2) Proximal Sensing (Hyperspectral Imagery) for Plant/Crop Biomass Quality/Quantity Modeling

(c) Hyperspectral Data Collection in the Field

(b) Imagery of a Plant

(3) UAV-based High Resolution Mapping Applications (feature extraction, target detection, etc.)

A demo of 3D model creation of a building before tearing down.

(a) Land Surface Model

(b) High Resolution RGB Imagery

(c) Elevation Point Cloud

(d) Flight Preparation

(e) A Snap-shot of the UAV During Flight

Last updated: July 5, 2022.