You are here:
Land-cover changes to surface-water buffers in the Midwestern USA: 25 years of Landsat analyses (1993-2017) - ICRW7 presentation
Citation:
Lane, C., T. Berhane, S. Mengistu, J. Christensen, H. Golden, S. Qiu, Z. Zhu, AND Q. Wu. Land-cover changes to surface-water buffers in the Midwestern USA: 25 years of Landsat analyses (1993-2017) - ICRW7 presentation. ICRW7 – Seventh Interagency Conference on Research in the Watershed, Virtual, November 16 - 19, 2020.
Impact/Purpose:
Changes wrought to areas buffering waters likely affect the processes within a given water as well as cumulatively modify downstream conditions, and changes are being wrought to smaller waters at faster rates than large systems.
Description:
Important biogeochemical processing and flow-attenuation occurs in the surface and near-surface inflows surrounding - or buffering - aquatic systems. Hence, modifications to the areas buffering waters can have profound impacts on the quantity, quality, and seasonal inundation in a given water body, as well as implications for the condition of downstream systems. To understand the timing, extent, magnitude, and frequency of change in the buffer areas (90-meter) surrounding waters of the Midwestern US, we analyzed the full archive of three Landsat path/row combinations totaling 31 years of data for ~100,000 km2, including areas of high urbanization (i.e., Chicago, IL, and St. Louis, MO) as well as agriculturally dominated landscapes (i.e., Peoria, IL). We used the Continuous Change Detection and Classification (CCDC) algorithm, which identified instances of land-use/land-cover (LULC) change throughout the continuous Landsat archive for each 30-m pixel. We trained a random forest classification algorithm using the 2001 National Land Cover Dataset (NLCD), binned the data into six LULC classes, and analyzed continuous LULC change with CCDC from 1993-2017. Though relatively small as a percent of the image, the spatial extent of the LULC modifications is substantial (e.g., developed lands increased by 280 km2 in the Chicago image, whereas ~300 km2 of forested land was converted to other LULCs in the St. Louis image). While change was consistent for the ~110,000 waterbody buffers analyzed across all three images, LULC in the buffering areas frequently occurred at much greater rates than LULC calculated image-wide. For instance, buffer LULC changed to developed lands at 2x the rate in the Chicago image and 3x in Peoria. Forested and grasslands buffering waters were converted to other LULCs at 7x and 3x, respectively, the image-wide rate of change in the Chicago image. However, not all LULC change was conversion to development or agricultural classes, as waterbody expansion in the buffers occurred at rates of 13-70x the image-wide rates across the three images. The greatest change occurred most frequently in the buffers surrounding the smallest waters (<0.1 ha), though this varied by image. Changes wrought to areas buffering waters likely affect the processes within a given water as well as cumulatively modify downstream conditions. Therefore, incorporating water buffer LULC dynamics into large-scale modeling and empirical studies will improve the physical representation of the landscape and affect research on aquatic nutrient and hydrologic dynamics.
URLs/Downloads:
https://www.icrwatersheds.org/seventh-interagency-conference-on-research-in-the-watersheds/