ColoradoView Student Interns Research
The project retrieves land surface temperature (LST) using a single-window algorithm on Landsat 8 data between 2014 and 2018 and investigated the land surface temperature (LST) pattern in Colorado between 2000 and 2018 using USGS Analysis Ready Data product. All data were processed on Google Earth Engine. The retrieved LST maps were validated against the ARD LST product. The difference of ARD LST distributions between the three Landsat platforms were analyzed. The long-term LST pattern by land use types (i.e., six dominant ones in Colorado, developed, barren, forest, shrubland, grassland and pasture, and crops) were also analyzed.
Interns: Eric Jensen, Mariah Wang, Nicolas Vogel
U.S. Landsat Analysis Ready Data (ARD) recently included the Land Surface Temperature (LST) product, which contains widespread and irregularly-shaped missing pixels due to cloud contamination or incomplete satellite coverage. Many analyses rely on complete LST images therefore techniques that accurately fill data gaps are needed. Here, the development of a partial-convolution based model with the U-Net like architecture to reconstruct the missing pixels in the ARD LST images is discussed.
Intern: Ben Newell
The Invasive Species research project is currently making use of plot data created by the Bureau of Land Management alongside MODIS fire map imagery to investigate the spread of B. tectorum(Cheat Grass) across the rangelands west of the Continental Divide. These investigations and resulting analyses will provide the basis for developing achievable management plans to stop the spread of cheatgrass further east.
Interns: Josh Otis, Matt Edrich, Catherine Jarnevich, Helen Sofaer
Reports: ColoradoView Invasive Species Report
The Pawnee National Grasslands are designated as an area of ecological concern in terms of Colorado’s endangered, indigenous species such as the Mountain Plover or Charadrius montanus. The Mountain Plover relies on a proportional amount of shortgrass prairie for nesting and bare soil for hunting. Fledgling success is shown to be related to drought conditions, which determine proportion and health of shortgrass species. In recent years, there has been a significant decline in shortgrass due to alternating as well as simultaneous effects of drought, changes in fire regimes, and over-grazing. By monitoring drought conditions, one can identify decreasing trends in NDVI when drought is not present which may indicate potential over-grazing. However, when drought is present, the effects of such can be analyzed by the severity of drought in comparison to the significance of the decrease in NDVI to determine any other potential influences on NDVI.
Interns: Jillian LaRoe, Joanitha Vogel
The overall goal of the UV-B project was to measure and geographically quantify the UV-B irradiance reaching the earth’s surface in the USA. There are 36 actual sites spread throughout the USA measuring UV-B radiation on the ground to assess its impact on agriculture. To enhance the data collected from these monitoring sites and create a better national model of UV-B radiation, it is being supplemented with satellite data from the Ozone Monitoring Instrument (OMI). Specifically, we used OMUVBd data, a data product measuring surface UV irradiance and erythemal dose based on ozone levels, cloud cover, and various light wavelengths reaching the surface and reflected back to the sensor.
Intern: Luke Martin
Report: ColoradoView UV-B Project
Three species of toadflax are considered to be invading Colorado: yellow toadflax (Linaria vulgaris), dalmation toadflax (Linaria dalmatica) and a hybrid of the two. Much is known about yellow and dalmation toadflax; however, little is published on hybrid toadflax. All species are prolific, spreading rapidly in common conditions and across a wide range in elevation. The hybrid appears to take on characteristics beneficial for reproduction from its parents enabling it to spread to a larger distribution. The purpose of this project is to create a habitat suitability map for the hybrid toadflax species in Colorado, basing the prediction on species distribution models created for dalmation and yellow toadflax. The models will incorporate remote sensing data as well as climatic and topographic variables.
Interns: Kevin McCartney, Jordan Lestina, Max Cook
We conducted vegetation classifications on the BX ranch and the neighboring State Land Board. This ranch is part of a Sustainable Grazing pilot and will determine baseline vegetation conditions on the ranch so that we can assess improvements in land condition due to management alterations. Using ArcGIS we mapped the land cover of the area, classifying areas of different vegetation health, using property boundaries provided from Teresa Chapman, USDA NAIP imagery downloaded from EarthExplorer, and a base map. We documented differences in vegetation classifications and tried to determine which patches have maintained health (greenness and cover) throughout the past decade of drought. Ideally, we will be able to use these methods to map the entire Southeastern shortgrass prairie of Colorado.
Interns: Kris Rogers, Alex Nelson
Report: ColoradoView Sustainable Grazing
The UV-B monitoring project focused on modeling the ultraviolet radiation index across the state. In Colorado, UV-B is measured directly at only 3 places and a map is prepared from readings at those stations. Because the number of stations is not comprehensive for the entire state, the maps are not representative of the correct UV-B radiation received by any point in the state of Colorado. Incoming UV-B radiation is dependent on a lot of factors, including but not limited to, solar irradiance, ozone layer cover, cloud cover, and the number of aerosols. Better maps are needed by farmers to estimate crop yield and reduce crop damage due to UV-B radiation. The main focus of this project was to gather relevant datasets that would aid in the modeling of incoming UV-B radiation.
Intern: Alfonso de Lara
Wheat is an economically important crop in Colorado with approximately 2.2 million acres planted in the state at total value of around $580 million based on 2012 harvest information. Wheat stem sawfly (Cephus cintus) is a significant pest of wheat in Colorado and other wheat growing areas of the United States and can have severe impacts on production. So far, no study has been conducted that has mapped the current and potential distribution of this species in Colorado. This project will integrate remote sensing imagery with known distributions of wheat stem sawfly to improve distribution model predictions. Wheat stem sawfly (WSS) presence data has been collected from Eastern Colorado from 2012 to 2014 by researchers in the Bioagricultural Sciences and Pest Management Department in the College of Agriculture at CSU and has been shared for the purpose of this study. The purpose of this study is to map the current and potential distribution of wheat stem sawfly in Colorado. This study will determine bioclimatic variables that are significantly correlated with wheat stem sawfly infestation and apply this information to predict future areas that this species can spread to in Colorado. This study integrates ecological theory with geospatial systems and statistical knowledge to achieve the proposed goals.
Interns: Jordan Lestina, Maxwell Cook
The grazing lands project for ColoradoView has aimed to fill gaps in the availability of spatial data on grazing lands in Colorado. Initial efforts to find any accurate, open source and comprehensive spatial data on grazing lands in Colorado proved fruitless. The main objective of this project was to utilize satellite imagery (Landsat and MODIS) to map grazing lands and grazing intensity in parts of Eastern Colorado. Interns aimed at calibrating MODIS and Landsat data with field data provided by the USDA. The USDA field data is separated into 3 different grazing plots with each respectively representing light, medium and heavy grazing intensities. In order to utilize this data, interns attempted to find and download corresponding data from MODIS that would be comparable to the ground data. However, MODIS data is at such a rough scale (250m – 1000m) that there is plenty of work to be done before MODIS satellite data can be compared to the ground data. This included transforming the MODIS NDVI data (250m) to fPAR using a formula provided by Dr. Gonzalo Irisarri, a USDA researcher.
Interns: Eric Rounds, Riley Ross
Report: Grazing Lands