Qinghai Lake

Qinghai Lake is the page’s accepted English name; scientific publications also use Lake Qinghai. The Chinese romanization Qinghai Hu and historical English form Koko Nor refer to the same water body, not to Qinghai Province or to the entire catchment.^[7] A Chinese geospatial record places the water between 36°32′56″ and 37°14′05″N and 99°36′58″ and 100°45′32″E; a 2019 synthesis rounds those limits slightly.^[1][2]

The Datong Mountains border the basin on the north, the Riyue Mountains stand along the eastern divide, and the Qinghai Nanshan rise south of the lake. These uplands enclose a catchment of about 29,660 km². The lake surface is commonly reported near 3,194 m above sea level, but the cited synthesis does not identify the vertical datum, so that elevation should not be combined uncritically with values from satellite altimetry or other maps.^[2]

The 4,497 km² area used here is explicitly dated to 2017; it is not a timeless size. The lake’s roughly west–east long axis is about 106 km and its maximum width about 67 km in the same synthesis. A separate geospatial dataset mapped 4,304.79 km² and 501.04 km of shoreline in 2015. Because those products use different dates and may trace islands, marsh edges, and shallow marginal water differently, their figures describe particular mapped shorelines rather than an error-free permanent boundary.^[1][2]

Depth figures require a similar caution. The traceable bathymetry is an Academia Sinica survey from 1961–63: its maximum was 28.7 m, while volume divided by area gives an 18.43 m mean. An FAO analysis adjusted that chart to the lower 1990 lake surface and estimated a 26.9 m maximum and 19.03 m mean; the method used planimetry of the published contours and assumed an unchanged deepest-bed elevation of 3,167.3 m.^[3]

Recent papers commonly repeat a 21 m mean and a 32–32.8 m maximum, but do not link those numbers to a newer bathymetric survey or state a sounding datum. This page therefore reports both families of values and uses the documented survey in the stat card. A changing surface level can alter water depth, but by itself it cannot explain a deeper modern maximum than the older chart.^[2][3]

Qinghai Lake is not simply a hollow between mountains. A structural study by the Chinese Academy of Sciences interprets Pliocene fault subsidence beneath the basin and formation of the lake during the Middle Pleistocene. It divides the lake basin into three horsts—blocks raised relative to adjacent crust—and three secondary fault depressions. The authors relate those structures to reactivation along the southern margin of the Middle Qilian block, the Zhongwulongshan–Qinghai Nanshan belt, the Heimahe–Dari belt, and cross-cutting faults of several orientations.^[4]

This is a tectonic interpretation, not a precise birth date for the present shoreline. Faulting and differential uplift supplied the structural accommodation; erosion from the surrounding ranges and deposition by rivers and waves then modified the basin floor and margins. The resulting lake is wide relative to its depth and includes gently sloping shelves, delta fronts, local bedrock highs, and several structural lows rather than one narrow rift trench.^[4][6]

The Buha and Wuha’alan systems build deltaic deposits on the northwest margin, while the Shaliu forms a fan delta on the northeast. Along the south, sediment arrives from short alluvial fans descending from the Qinghai Nanshan. On the southeast margin, material from the Riyue uplands mixes with windblown sand. Field mapping and remote sensing distinguish bedrock-derived, alluvial-fan, and braided-river beach-bar systems around these shores.^[6]

Beach bars are ridges of sand or gravel reworked by waves. At Qinghai Lake they are especially developed on slopes of roughly 0.1°–0.7°, where a small vertical level change can move the waterline far across the shore. Bars run parallel to parts of the northeast, southeast, and south coasts and partly or completely enclose lagoons. Prevailing west and northwest winds set up waves and currents that sort incoming sediment; changing lake level shifts the zone in which that reworking occurs.^[6]

There is no surface outlet. Six large inflows identified in a 1979–2018 coupled lake–watershed model are the Buha, Shaliu, Quanji, Haergai, Daotang, and Heima rivers. Together they supplied a modelled long-term mean of 1.594 × 10⁹ m³ a year, 66.8% of the model’s total river inflow of 2.348 × 10⁹ m³ a year. The Buha alone accounted for 46.7%. These are model-period averages, not measured discharge for a recent year.^[5]

Expressed as equivalent depth over the lake, that model’s 1979–2018 annual means were 386 mm of precipitation, 996 mm of evaporation, and 587 mm of river inflow. Direct runoff and groundwater crossing the modelled shoreline added 17 and 28 mm respectively; most land-surface and subsurface water first entered a river and was counted there instead. The partition differs from older water budgets, and the authors explicitly identify uncertainty in groundwater and total runoff.^[5]

Because evaporation removes water but leaves most dissolved ions behind, the outlet-free basin is brackish and alkaline. A widely used synthesis value is about 12.5 g/L salinity with pH 9.3, derived from earlier observations rather than a single current whole-lake survey. The FAO field programme found 12.4–12.5 g/L in South Bay profiles in September 1992 and documented higher lake-wide values in the mid-1980s, showing why salinity should be tied to sampling date and place.^[2][3]

At lake level, a 1959–2015 synthesis gives a mean annual air temperature of 1.9°C and about 340 mm of annual precipitation, more than 65% of it in summer. A later basin-wide reanalysis gives −2.5°C and 415 mm for 1919–2018. The difference is largely one of domain and dataset: a high mountain catchment average is not interchangeable with conditions at the lake shore. Summer monsoon moisture supplies most precipitation, while strong winds, dry continental air, and available heat drive evaporation from the open water.^[2][5]

Ice commonly covers the lake from December or January until early April. A thermodynamic study classifies Qinghai Lake as dimictic: it mixes through much of the water column twice each year, around the spring and autumn transitions, while summer heating and winter inverse stratification produce intervening layered periods. The timing of freeze-up and break-up varies from year to year.^[2][3]

Lake-level change has not followed one continuous trend. A synthesis reports an average fall of 7.6 cm a year from 1961 to 2004, when evaporation generally exceeded river inflow plus precipitation, followed by an average rise of 14 cm a year during 2004–2012 as runoff and precipitation increased. A separate 1979–2018 coupled model reproduced the observed decrease-then-increase pattern. These rates describe their stated intervals and should not be projected as permanent trends.^[2][5]

The Riyue Mountains and Qinghai Nanshan separate the Qinghai Lake catchment from neighboring outward drainage, including parts of the upper Yellow River system. Water falling inside the Qinghai Lake divide terminates in the lake; water beyond the eastern divide can ultimately join the Yellow River. A 2021 geological synthesis interprets uplift along the Riyue–Yeniu margin as having severed a former eastward connection during the Middle to Late Pleistocene, but that interpretation does not establish an exact closure date.^[8] The hydrologic boundary, rather than the administrative limits of Qinghai Province, defines the basin described here.

For related physical geography, use the lake hub, mountain hub, river hub, nearby but hydrologically separate Yellow River record, and the terrain index entries for faults, deltas, beach ridges, dunes, and drainage divides.

1. Gou, Z., Liu, F., Liu, C., and Chen, L., Outline of Qinghai Lake, Digital Journal of Global Change Data Repository (March 2018; accessed 30 August 2026). Geospatial record produced by Qinghai Normal University, the Chinese Academy of Sciences’ Institute of Geographic Sciences and Natural Resources Research, and the National Geomatics Center of China. Source for name forms, coordinate bounds, surrounding ranges, mapped 2015 area and shoreline, and named inflows.
2. Su, D. et al., “Numerical study on the response of the largest lake in China to climate change”, Hydrology and Earth System Sciences 23, 2093–2109 (2019; accessed 30 August 2026). Source for the 2017 area, 29,660 km² catchment, coordinate range, dimensions, commonly cited elevation and depth values, salinity and pH synthesis, 1959–2015 climate normals, ice season, mixing regime, and interval-specific level trends. The paper does not state a vertical datum or identify a new bathymetric survey for the 32.8 m maximum.
3. Walker, K. F., A Management Plan for the Naked Carp Fishery of Qinghai Lake, chapter 3, “The Lake Environment”, UNDP/FAO project CPR/88/077 (January 1993; accessed 30 August 2026). Source for the 1961–63 survey bathymetry, the documented 1990 planimetric adjustment, area–volume caveats, measured 1990–92 salinity and pH, and seasonal water-column structure. The report reproduces older measurements and clearly identifies where later values are calculated rather than sounded.
4. Bian, Q., Liu, J., Luo, X., and Xiao, J., “Geotectonic Setting, Formation and Evolution of the Qinghai Lake”, Seismology and Geology 22(1), 20–26 (2000; accessed 30 August 2026). Source for the Pliocene fault-depression interpretation, Middle Pleistocene lake formation, three horsts and three subsidiary depressions, named fault belts, and differential uplift. These are the authors’ structural interpretations, not a directly dated age for the modern shoreline.
5. Shu, L. et al., “Advancing understanding of lake–watershed hydrology: a fully coupled numerical model illustrated by Qinghai Lake”, Hydrology and Earth System Sciences 28, 1477–1491 (2024; accessed 30 August 2026). Source for the six principal inflows, 1979–2018 modelled runoff shares and water-budget components, observational period, level reversal, and stated disagreements among runoff and groundwater estimates. All quoted fluxes are model-period means.
6. Peng, B., Xu, Z., Chen, X., and Zhao, Y., “Distribution patterns and controlling factors of beach bars in Qinghai Lake, NW China”, Frontiers in Earth Science 12, article 1390039 (2024; accessed 30 August 2026). Source for delta-to-beach sediment pathways, mapped beach-bar and lagoon distribution, wind and wave controls, shore and basin slopes, and bedrock, fan-delta, and braided-river depositional systems.
7. Library of Congress, Library of Congress Subject Headings, Section Q, entry “Qinghai Lake (China),” p. Q-4 (archived 34th edition; accessed 30 August 2026). Name authority for Qinghai Lake, with Qinghai Hu, Koko Nor, and Kuku Nor recorded as variant forms.
8. Tao, H., Hao, L., Li, S., Wu, T., Qin, Z., and Qiu, J., “Geochemistry and Petrography of the Sediments From the Marginal Areas of Qinghai Lake, Northern Tibet Plateau, China”, Frontiers in Earth Science 9, article 725553 (2021; accessed 30 August 2026). Source for the interpreted former Paleo-Buha–Daotang–Yellow River connection, Middle-to-Late-Pleistocene separation at the eastern mountain margin, and the distinction between modern closed drainage and neighboring outward drainage. The closure history is a synthesis of earlier work, not a new direct date.