Low Water Episodes
133 low water episodes at Memphis, St. Louis, Vicksburg: how often, how long, by decade and what happened after earlier ones.
Informational only, not for navigation. Read the notice.
Key Facts
- At Memphis, 29 separate episodes below -8.00 ft are recorded from 1988 to 2025, in 127 years of archive: the median lasted 7 days and the longest 54 days, from 9 September 2023 to 1 November 2023.
- At St. Louis, 25 separate episodes below -3.20 ft are recorded from 1961 to 2025, in 67 years of archive: the median lasted 4 days and the longest 39 days, from 15 December 1963 to 22 January 1964.
- At Vicksburg, 79 separate episodes below 0.80 ft are recorded from 1901 to 2024, in 126 years of archive: the median lasted 8 days and the longest 166 days, from 7 September 1939 to 19 February 1940.
- No gauge on this page is in a low water episode on the data date.
Memphis
127 years of archiveMemphis gauge page. A line is a published low water level; an episode is an unbroken run of days at or past it.
| Line | Episodes | Per year of record | Median length | Longest | Over 30 days |
|---|---|---|---|---|---|
| NWS low water threshold: below -8.00 ft | 29 | 0.23 | 7 d | 54 d 2023-09-09 to 2023-11-01 | 7% |
By decade: NWS low water threshold
| Decade | Years of data | Episodes | Per year | Days in episodes | Longest |
|---|---|---|---|---|---|
| 1900s | 10 | 0 | 0.00 | 0 | 0 d |
| 1910s | 10 | 0 | 0.00 | 0 | 0 d |
| 1920s | 10 | 0 | 0.00 | 0 | 0 d |
| 1930s | 10 | 0 | 0.00 | 0 | 0 d |
| 1940s | 10 | 0 | 0.00 | 0 | 0 d |
| 1950s | 10 | 0 | 0.00 | 0 | 0 d |
| 1960s | 10 | 0 | 0.00 | 0 | 0 d |
| 1970s | 10 | 0 | 0.00 | 0 | 0 d |
| 1980s | 10 | 7 | 0.70 | 97 | 36 d |
| 1990s | 10 | 0 | 0.00 | 0 | 0 d |
| 2000s | 10 | 7 | 0.70 | 31 | 13 d |
| 2010s | 10 | 5 | 0.50 | 43 | 20 d |
| 2020s | 7 | 10 | 1.43 | 140 | 54 d |
Longest and most recent
| Kind | From | To | Days | Lowest daily mean |
|---|---|---|---|---|
| Longest | 2023-09-09 | 2023-11-01 | 54 | -11.94 ft on 2023-10-17 |
| 2nd longest | 1988-06-18 | 1988-07-23 | 36 | -10.70 ft on 1988-07-10 |
| 3rd longest | 1988-08-26 | 1988-09-19 | 25 | -10.10 ft on 1988-09-16 |
| 4th longest | 2022-10-12 | 2022-11-04 | 24 | -10.79 ft on 2022-10-22 |
| 5th longest | 2024-10-18 | 2024-11-07 | 21 | -10.31 ft on 2024-11-03 |
| Most recent | 2025-10-25 | 2025-10-28 | 4 | -8.27 ft on 2025-10-26 |
| Recent 2 | 2025-10-17 | 2025-10-23 | 7 | -8.86 ft on 2025-10-20 |
| Recent 3 | 2024-10-18 | 2024-11-07 | 21 | -10.31 ft on 2024-11-03 |
| Recent 4 | 2024-09-21 | 2024-09-27 | 7 | -10.27 ft on 2024-09-24 |
| Recent 5 | 2023-12-23 | 2023-12-30 | 8 | -9.15 ft on 2023-12-26 |
St. Louis
67 years of archiveSt. Louis gauge page. A line is a published low water level; an episode is an unbroken run of days at or past it.
| Line | Episodes | Per year of record | Median length | Longest | Over 30 days |
|---|---|---|---|---|---|
| NWS low water threshold: below -3.20 ft | 25 | 0.37 | 4 d | 39 d 1963-12-15 to 1964-01-22 | 4% |
By decade: NWS low water threshold
| Decade | Years of data | Episodes | Per year | Days in episodes | Longest |
|---|---|---|---|---|---|
| 1960s | 10 | 12 | 1.20 | 109 | 39 d |
| 1970s | 10 | 2 | 0.20 | 6 | 4 d |
| 1980s | 10 | 2 | 0.20 | 25 | 21 d |
| 1990s | 10 | 0 | 0.00 | 0 | 0 d |
| 2000s | 10 | 3 | 0.30 | 11 | 4 d |
| 2010s | 10 | 4 | 0.40 | 27 | 13 d |
| 2020s | 7 | 2 | 0.29 | 3 | 2 d |
Longest and most recent
| Kind | From | To | Days | Lowest daily mean |
|---|---|---|---|---|
| Longest | 1963-12-15 | 1964-01-22 | 39 | -5.54 ft on 1963-12-31 |
| 2nd longest | 1989-12-12 | 1990-01-01 | 21 | -5.12 ft on 1989-12-23 |
| 3rd longest | 1963-01-19 | 1963-02-08 | 21 | -5.72 ft on 1963-01-26 |
| 4th longest | 1961-01-29 | 1961-02-13 | 16 | -4.06 ft on 1961-01-30 |
| 5th longest | 2012-12-28 | 2013-01-09 | 13 | -4.38 ft on 2012-12-31 |
| Most recent | 2025-02-19 | 2025-02-19 | 1 | -3.35 ft on 2025-02-19 |
| Recent 2 | 2025-01-20 | 2025-01-21 | 2 | -3.38 ft on 2025-01-20 |
| Recent 3 | 2013-01-22 | 2013-01-28 | 7 | -4.16 ft on 2013-01-24 |
| Recent 4 | 2012-12-28 | 2013-01-09 | 13 | -4.38 ft on 2012-12-31 |
| Recent 5 | 2012-12-15 | 2012-12-17 | 3 | -3.45 ft on 2012-12-16 |
Vicksburg
126 years of archiveVicksburg gauge page. A line is a published low water level; an episode is an unbroken run of days at or past it.
| Line | Episodes | Per year of record | Median length | Longest | Over 30 days |
|---|---|---|---|---|---|
| NWS low water threshold: below 0.80 ft | 79 | 0.63 | 8 d | 166 d 1939-09-07 to 1940-02-19 | 16% |
By decade: NWS low water threshold
| Decade | Years of data | Episodes | Per year | Days in episodes | Longest |
|---|---|---|---|---|---|
| 1900s | 9 | 4 | 0.44 | 62 | 44 d |
| 1910s | 10 | 2 | 0.20 | 21 | 16 d |
| 1920s | 10 | 0 | 0.00 | 0 | 0 d |
| 1930s | 10 | 8 | 0.80 | 295 | 166 d |
| 1940s | 10 | 12 | 1.20 | 272 | 74 d |
| 1950s | 10 | 17 | 1.70 | 458 | 126 d |
| 1960s | 10 | 12 | 1.20 | 202 | 128 d |
| 1970s | 10 | 3 | 0.30 | 25 | 13 d |
| 1980s | 10 | 8 | 0.80 | 126 | 48 d |
| 1990s | 10 | 1 | 0.10 | 4 | 4 d |
| 2000s | 10 | 3 | 0.30 | 13 | 6 d |
| 2010s | 10 | 4 | 0.40 | 34 | 21 d |
| 2020s | 7 | 5 | 0.71 | 35 | 17 d |
Longest and most recent
| Kind | From | To | Days | Lowest daily mean |
|---|---|---|---|---|
| Longest | 1939-09-07 | 1940-02-19 | 166 | -7.00 ft on 1940-02-03 |
| 2nd longest | 1963-09-09 | 1964-01-14 | 128 | -5.80 ft on 1964-01-06 |
| 3rd longest | 1953-09-16 | 1954-01-19 | 126 | -4.40 ft on 1953-11-27 |
| 4th longest | 1956-09-30 | 1956-12-17 | 79 | -4.10 ft on 1956-10-28 |
| 5th longest | 1940-09-20 | 1940-12-02 | 74 | -6.80 ft on 1940-10-31 |
| Most recent | 2024-11-04 | 2024-11-05 | 2 | 0.38 ft on 2024-11-05 |
| Recent 2 | 2024-10-30 | 2024-11-01 | 3 | 0.47 ft on 2024-10-31 |
| Recent 3 | 2023-10-12 | 2023-10-28 | 17 | -1.37 ft on 2023-10-20 |
| Recent 4 | 2023-10-01 | 2023-10-06 | 6 | 0.44 ft on 2023-10-03 |
| Recent 5 | 2022-10-21 | 2022-10-27 | 7 | -0.39 ft on 2022-10-25 |
How This Is Calculated, and What It Cannot Show
An episode is a maximal run of consecutive calendar days whose daily mean is on the stated side of a published line. A day with no reading ends a run, so a gap in the record can split one long episode into two. The line is the one shown on the gauge page and cited there. When an archive starts already past a line, that first run is left out because its true start is unknown, and a run still going is kept apart so every statistic is about episodes whose whole length is known.
A gauge appears only with at least 10 years of daily means, and a line only with at least 3 completed episodes. "How earlier episodes ended" compares a current episode with earlier ones that lasted at least as long, needs at least 5 of them, and counts how many ended within 14 more days. It describes the past record and is not a forecast: river conditions change, and the archive is one realisation of them. Values before the authority validated them are raw, and these are our own calculations, not the authority's.
A century of stage readings is not a like for like record. A riverbed can scour or fill, a channel can be dredged or trained, and an authority can move a gauge or change its datum, so "below a fixed number" has not always meant the same drought. The decade table therefore compares how often the stage fell below one fixed number, not how severe each drought was.
The Rhine, Danube and Elbe gauges are not here because their archives are too short to hold even 3 completed episodes. They will appear when the record is long enough.
Source: NOAA National Water Prediction Service (water.noaa.gov). Gauge data from USACE and USGS. (US government work) Data as of 2026-11-20 (20 November 2026).