Full text
Documentation for the coordinated Lake Superior lakewide average monthly mean water levels (1918-2024) IGLD 1985 Last updated: October 2025 Title Lake Superior lake-wide average monthly mean water level Source Coordinated using data computed by Environment and Climate Change Canada (ECCC) and the United States Army Corps of Engineers (USACE) from hydrometric observations collected by the Canadian Hydrographic Service (CHS) and the National Oceanic and Atmospheric Administration (NOAA). Institution The Coordinating Committee on Great Lakes Basic Hydraulic and Hydrologic Data (Coordinating Committee, for short) Summary The Lake Superior lake-wide average monthly mean water level is computed by ECCC and USACE from daily mean water level observations collected from gauges located around the lake, and subsequently coordinated by the Coordinating Committee. The data are in meters with reference to IGLD 1985. Detailed Description The monthly mean lake-wide average water level for Lake Superior is computed by USACE and ECCC using hydrometric observations from five gauges located around the lake. Gauge locations have been selected based on their data quality, period and record, and their location around the lake to best capture the average water level for the entire lake surface. In other words, water level measured at a single gauge may not be representative of the level across the lake surface owing to local rises or falls in water level due to meteorological factors (e.g., wind setup). The names, agency, and gauge ID for the five gauges are: • Duluth, MN, NOAA, 9099064 • Marquette C.G., MI, NOAA, 9099018 • Point Iroquois, MI, NOAA, 9099004 • Thunder Bay, CHS, 10050 • Michipicoten, CHS, 10750 The method for computing the monthly mean lake-wide average water level is as follows: 1. Daily mean water level data for each gauge is collected from CHS and NOAA. The water levels are provided to the nearest millimeter with respect to IGLD 1985. 2. The daily mean water levels from each gauge are then averaged to the nearest centimeter using engineering rounding*.
3. The rounded daily means from each of the five gauges are then averaged to the nearest centimeter using engineering rounding to compute the lake-wide average daily mean water level†. 4. The lake-wide average daily mean water levels are then averaged to the nearest centimeter using engineering rounding to compute the lake-wide average monthly mean water level. 5. Each agency, ECCC and USACE, follows the above procedure to compute the lake-wide average monthly mean water level. Annually, the Coordinating Committee will compare the most recent year of data to verify that their calculations were consistent. The annual verification process uses finalized data from CHS and NOAA that at times may have been updated from the provisional data available in near real time. *In this dataset, levels are rounded to the thousandth decimal place. Engineering rounding, or rounding half to even, is used to avoid the bias given to larger values due to conventional rounding (rounding half up). For example, the value of 176.555 would be rounded to 176.56 in both conventional rounding and engineering rounding. The value of 176.565 would be rounded to 176.56 using engineering rounding and 176.57 using conventional rounding. The key digit in these examples is the thousandth decimal place. If the thousandth decimal place is 0-4, no rounding is performed on the value in the hundredth decimal place. For values 6-9 in the thousandth decimal place, the value in the hundredth place is rounded up. If the thousandth place is a 5 the value for the hundredth place is always rounded to the even value. †If data is missing from a gauge for a given day, then the data from that gauge is excluded from the lake-wide average daily mean water level calculation. In addition, only a subset of the five gauges is used to compute the lake-wide average daily mean water level. The logic used to determine the gauge subset in the event of missing data is: • Missing Point Iroquois: average Michipicoten, Duluth and Marquette • Missing Point Iroquois and Michipicoten: average Thunder Bay, and Marquette • Missing Michipicoten: average Point Iroquois, Thunder Bay, and Marquette • Missing Duluth: average Point Iroquois, Thunder Bay, and Marquette • Missing Thunder Bay: Average Point Iroquois, Michipicoten, and Duluth • Missing Marquette: average Point Iroquois, Michipicoten, and Duluth • Missing 2 or more: Average remaining gauges Revision History The Coordinating Committee first published this dataset on its website in 2017. Notes CHS publishes this dataset along with associated long-term statistics at the following website: https://www.tides.gc.ca/en/monthly-historical-great-lakesmeans
USACE publishes the long-term statistics and hydrographs at the following website: https://water.usace.army.mil/office/lre/waterleveldata/waterLevelData These data are available starting in 1918 because before that time, there were too few gauges to compute a reasonable lake-wide average. Starting around 1860, each lake had one water level station that was designated as the “master gauge” based on length of record and minimal vertical crustal movement. Over the years, the gauges included in the network for each lake has been expanded and modified. As a result, there is some uncertainty regarding the accuracy of some older data due to changes in the gauge network over time. Historic Gauge Network Variations Missing Gauges Time Periods Gauges Used for Computation Point Iroquois, MI Jan 1918 to Oct 1930 Mar 1931 Oct 1944 to Oct 1950 Thunder Bay Marquette G.C., MI Michipicoten Duluth, MN Point Iroquois, MI Michipicoten Mar 1918 Dec 1919 Feb 1920 Sept 1920 Apr 1924 Feb 1925 Mar 1925 Apr 1925 Thunder Bay Marquette G.C., MI Michipicoten Apr 1966 Point Iroquois, MI Thunder Bay Marquette G.C., MI Thunder Bay Feb 1978 Point Iroquois, MI Michipicoten Duluth, MN
