Introduction
Earth’s hydrosphere is overwhelmingly saline: oceans and marginal seas, together with saline groundwater and saline endorheic lakes, constitute over 97% of planetary water. Fresh water represents only a small fraction of the total—on the order of 1%—so that the mass ratio of salt water to fresh water is roughly 50:1. Mean ocean salinity is about 35‰ (≈3.5% by mass, roughly 34–35 g dissolved salts per kilogram of seawater), although local values vary with the amount of terrestrial runoff and other regional processes.
Salinity thresholds are commonly used to distinguish water types for human and ecological purposes. Freshwater is typically defined as having salinity below about 0.35‰ (approximately 1% of mean ocean salinity); waters with salinity between ~0.35‰ and 1‰ are often termed marginal because their suitability for many uses is borderline. The largest reservoirs of contemporary fresh water are cryospheric and subsurface: most fresh water is stored as ice and snow, as groundwater, and as soil moisture. Only a very small fraction of the planet’s fresh water—about 0.3% of the freshwater budget—exists as liquid water at the surface.
Read more Government Exam Guru
Within that limited surface-liquid pool, lakes hold by far the greatest share (≈87%), wetlands (swamps and marshes) account for roughly 11%, and rivers contain only about 2%; still smaller amounts reside in the atmosphere and in biota. Groundwater volumes exceed annual river runoff, but a substantial portion of groundwater is saline and should be treated with saline reservoirs when evaluating usable freshwater supplies. In assessments of arid regions, saline aquifers are sometimes included because they can be locally significant for supply or contamination, yet many such aquifers contain palaeowater—fossil groundwater recharged millennia ago—that is effectively nonrenewable on human timescales and should not be considered sustainable.
The present-day partitioning of freshwater—dominated by ice, snow, and subsurface stores—contrasts with warmer intervals in Earth’s history (e.g., much of the Mesozoic and Paleogene), when glaciers were absent and most freshwater occurred as liquid flow in rivers and streams. This geological perspective underscores that contemporary freshwater distribution is dynamic and sensitive to climate and long-term hydrological change.
Distribution of saline and fresh water
Free Thousands of Mock Test for Any Exam
Earth’s hydrosphere comprises approximately 1.386 × 10^9 km3 of water, of which roughly 97.5% is saline and 2.5% is fresh. By volume the oceans are overwhelmingly dominant: their combined volume (~1.338 × 10^9 km3) constitutes about 96.5% of all water and accounts for roughly 99% of the planet’s salt water, covering ~70.8% of the surface and producing the planet’s characteristic blue appearance. The Pacific is the largest single reservoir (≈6.70 × 10^8 km3; ~48.3% of global water), followed by the Atlantic (≈3.10 × 10^8 km3; ~22.4%), the Indian (≈2.64 × 10^8 km3; ~19.0%), the Southern (≈7.18 × 10^7 km3; ~5.2%), and the Arctic (≈1.87 × 10^7 km3; ~1.4%).
Continental ice and surface snow (the cryosphere) and glaciers store a significant fraction of freshwater (≈2.4364 × 10^7 km3, ~1.76% of total water). Much of this is locked in large ice sheets: the Antarctic ice sheet alone contains ≈2.16 × 10^7 km3 (~1.56% of total water), the Greenland ice sheet ≈2.34 × 10^6 km3 (~0.17%), with the remainder in mountain glaciers, Arctic islands, and permafrost.
Groundwater constitutes another major freshwater reservoir (≈2.34 × 10^7 km3, ~1.69% of total water). That pool includes both saline groundwater (≈1.287 × 10^7 km3, ~0.93% of total) and fresh groundwater (≈1.053 × 10^7 km3, ~0.76% of total); fresh groundwater supplies about 30% of what is categorized as the liquid-surface-fresh-water equivalent in common tabulations.
Surface freshwater (the portion most directly available for ecosystems and human use) is extremely limited in proportion to total water. Lakes hold only ≈1.764 × 10^5 km3 (~0.013% of global water), split nearly equally between saline and fresh lakes. Fresh lakes (≈9.10 × 10^4 km3; ~0.0066% of global water) account for roughly 87% of the liquid-surface-fresh-water equivalent. Key fresh-lake contributors include the African Great Lakes (~3.007 × 10^4 km3), Lake Baikal (~2.3615 × 10^4 km3) and the North American Great Lakes (~2.2115 × 10^4 km3), which together constitute the majority of fresh-lake storage. The Caspian Sea dominates saline-lake volume (≈7.82 × 10^4 km3).
Smaller terrestrial and atmospheric reservoirs together contribute only minute fractions of global water: soil moisture (~1.65 × 10^4 km3), the atmosphere (~1.29 × 10^4 km3), and swamps (~1.147 × 10^4 km3) each represent well under 0.002% of total water, though swamps comprise a notable share (~11%) of the liquid-surface-fresh-water equivalent. Flowing water and biotic stores are correspondingly tiny: rivers contain ≈2.12 × 10^3 km3 (~0.00015% of total water) and living biomass holds ≈1.12 × 10^3 km3.
These proportions demonstrate the extreme concentration of water mass within the oceans and continental ice sheets, whereas immediately accessible liquid freshwater at Earth’s surface (lakes, rivers, swamps, soil moisture, and water in biota) represents only a very small fraction of total water. Because reservoir volumes span many orders of magnitude, logarithmic representations are typically employed to visualize their relative sizes.
Lakes
Earth’s lakes hold about 199,000 km3 of water, with lake basins disproportionately concentrated at high northern latitudes. This broad spatial pattern places most lacustrine storage away from the densest global population centers and reduces their direct role in human water supply relative to other freshwater reservoirs.
The North American Great Lakes constitute a notable exception: by volume they account for roughly 21% of the planet’s freshwater and therefore represent one of the largest accessible fresh‑water stores. The Great Lakes Basin is densely settled and economically significant, supporting more than 35 million people. Numerous major cities line the system’s shores, including Canadian cities such as Thunder Bay, St. Catharines, Hamilton, Toronto, Oshawa, and Kingston, and U.S. cities such as Detroit, Duluth, Milwaukee, Chicago, Gary, Cleveland, Buffalo, and Rochester.
Read more Government Exam Guru
Groundwater
Fresh groundwater is a vital freshwater reservoir in arid and semi‑arid regions, where its spatial pattern often parallels surface river networks but its subsurface location confers a key advantage: storage is far less susceptible to evaporative losses than open reservoirs or dams, making it particularly valuable in hot, dry climates. In settings with strongly seasonal or erratic precipitation—such as Yemen—rainfall pulses during the wet season commonly recharge aquifers and establish groundwater as the primary irrigation source; in this way the subsurface system functions as a buffer that captures intermittent inputs and sustains water supply through prolonged dry periods.
Quantifying groundwater recharge is intrinsically more complex than measuring surface runoff because subsurface processes (infiltration, vadose‑zone storage, delayed flowpaths and spatial heterogeneity of hydraulic properties) are difficult to observe directly. These measurement challenges create a practical management bias: where some surface water is available, it is often preferred because it is easier to quantify and control. Estimates of recharge for a given region can nonetheless vary markedly with method—hydrogeological models, tracer and isotope studies, water‑balance calculations and remote‑sensing proxies can yield divergent results—which produces substantial uncertainty in calculations of sustainable yield and in overall resource assessments.
Free Thousands of Mock Test for Any Exam
Over‑exploitation of groundwater, including withdrawal of non‑renewable or fossil aquifers beyond natural recharge rates, is widespread and carries long‑term consequences. The Ogallala Aquifer illustrates how rapid early development without adequate consideration of recharge limits can lead to progressive drawdown, declining well yields and growing sustainability risks for agricultural and urban systems. These outcomes underscore the necessity of cautious development strategies and improved recharge estimation to inform sustainable groundwater management.
Distribution of river water
At any given time rivers contain only a small fraction of the planet’s freshwater—about 2,120 km3 (≈0.49% of surface freshwater). Because river systems are highly dynamic, hydrologists characterize them by flow (annual surface runoff) rather than static volume. Runoff, however, is extremely uneven in space.
Regionally, annual river runoff is concentrated in a few continents: Asia (excluding the Middle East) supplies ~13,300 km3 (30.6% of the world total), South America ~12,000 km3 (27.6%), North America ~7,800 km3 (17.9%), Oceania ~6,500 km3 (14.9%), Sub‑Saharan Africa ~4,000 km3 (9.2%), Europe ~2,900 km3 (6.7%), Australia ~440 km3 (1.0%), and the Middle East and North Africa ~140 km3 (0.3%). These aggregates mask sharp internal contrasts: sizable renewable resources can be tightly clustered while large areas remain water‑poor (for example, roughly one quarter of Australia’s limited renewable freshwater is concentrated in the sparsely inhabited Cape York Peninsula; conversely, jurisdictions such as Texas and South Africa have very low renewable supplies—Texas about 26 km3/year over ~695,622 km2; South Africa about 44 km3/year for ~1,221,037 km2).
A small number of basins dominate global runoff. The Amazon and Orinoco together produce some 6,500 km3/year—about 15% of global river discharge. East and South/Southeast Asia also contribute major shares: the Yangtze alone is on the order of 1,000 km3/year, and the broader South and Southeast Asian system yields roughly 8,000 km3/year (≈18% of the global total), including the Ganges (~900 km3/year), the Irrawaddy (~500 km3/year) and the Mekong (~450 km3/year). High‑latitude basins are likewise significant: Siberian systems (notably the Yenisey, the Ob at >500 km3/year and the Lena at >450 km3/year) account for very large freshwater volumes—the Yenisey basin alone holds more than 5% of the world’s freshwater at basin scale and is second only to the Amazon. Canada contains over 10% of global river runoff (in addition to large freshwater lake stores), with major contributors such as the Mackenzie (>250 km3/year) and the Yukon (>150 km3/year). Smaller areas can also exhibit intense runoff density; for example, New Guinea’s Fly and Sepik basins together yield over 300 km3/year within roughly 150,000 km2 of drainage area.
In sum, global river runoff is dominated by a handful of large tropical monsoon and high‑latitude basins, while arid and semi‑arid regions receive a disproportionately small share. This pronounced spatial imbalance has direct implications for water-resource planning, allocation and transboundary hydrology.
Area, volume, and depth of oceans
The oceans and seas listed together cover about 361 million km2 and contain approximately 1,370 million km3 of water, yielding an overall mean depth of 3,796 m. The Pacific Ocean is the largest and deepest of the major basins, with an area of 165.2 million km2, a volume of 707.6 million km3, and a mean depth of 4,282 m. The Atlantic and Indian Oceans are substantially smaller: the Atlantic has 82.4 million km2 and 323.6 million km3 with a mean depth of 3,926 m, while the Indian comprises 73.4 million km2 and 291.0 million km3 with a mean depth of 3,963 m. Mean depths among these oceans vary modestly (≈3,926–4,282 m), with the Pacific dominating both areal extent and stored water.
These depth and basin patterns reflect the nature of oceanic lithosphere, which is comparatively young, thin and dense; seafloor rocks largely formed after the breakup of Pangaea, indicating continual seafloor generation and recycling. The higher density of oceanic crust produces topographic depressions that are preferentially occupied by seawater; on terrestrial bodies without surface oceans, equivalent dense basaltic provinces instead appear as broad plains and elevated plateaux rather than filled basins.
Read more Government Exam Guru
The chemical character of seawater arises from continental geology and the hydrologic cycle. Low‑density continental rocks contain abundant, readily weathered salts of alkali and alkaline‑earth metals; weathering and fluvial transport deliver these dissolved ions to the oceans. Because evaporation returns freshwater to land but leaves dissolved salts behind, marine basins have accumulated these ions over geologic time, producing the persistent salinity of the oceans.
Variability of water availability
Temporal and spatial variability in water supply is a primary determinant of both freshwater ecosystem structure and human water security. Water that appears only during infrequent extreme wet years cannot be treated as a renewable source for sustained use; conversely, because most global runoff originates in regions with relatively stable climate, aggregate global runoff tends to exhibit low variability.
Free Thousands of Mock Test for Any Exam
High‑elevation water stores—glaciers and seasonal snow—provide comparatively dependable summer discharge, even where surrounding lowlands are arid. This seasonally concentrated meltwater synchronizes supply with peak agricultural demand and has historically supported dense human settlement and intensive irrigation (for example, ancient river‑basin civilizations and contemporary irrigated regions such as California’s San Joaquin Valley).
Australia and southern Africa constitute a major regional exception to typical climate–runoff relationships. Rivers there, across both temperate (Köppen C) and arid (Köppen B) classes, commonly show coefficients of variation in runoff up to three times greater than similarly classified rivers on other continents. This anomaly is rooted less in present climate than in long‑term geologic and pedologic history: soils in these continents are extremely old and strongly depleted of plant‑available nutrients, especially phosphorus. Native flora have evolved dense, highly prolific root systems (e.g., proteoid roots) to mine scarce nutrients; these root systems also consume a disproportionately large fraction of incoming precipitation, so sustained runoff typically does not begin until cumulative rainfall is large (on the order of several hundred millimetres).
The consequence is strikingly low runoff ratios in Australian and southern African catchments compared with global counterparts (examples by Köppen class): BWh (mean annual rainfall ≈250 mm) runoff ≈1% versus ≈10% elsewhere; BSh (≈350 mm) ≈3% versus ≈20%; Csa (≈500 mm) ≈5% versus ≈35%; Caf (≈900 mm) ≈15% versus ≈45%; Cb (≈1,100 mm) ≈25% versus ≈70%. These reduced runoff efficiencies make flow regulation by storage uneconomic in many basins: reservoir evaporation and transmission losses mean extraordinarily large impoundments would be required to yield modest reliable yields (Lake Eyre Basin provides a clear example), and in typical cases storage volumes must be many times larger to secure a fraction of the supply attainable in comparable catchments elsewhere.
Ecologically, high hydrological variability favors life‑history strategies adapted to episodic abundance and prolonged scarcity—species that reproduce rapidly during floods and endure extended dry intervals. An important exception to the low runoff pattern occurs in tropical (Köppen A) regions of Australia and southern Africa: despite poor mineral soils, tropical vegetation can exploit organic phosphorus and phosphorus delivered in rainfall, so tropical rivers there do not show the markedly depressed runoff ratios found in cooler or drier zones. Finally, regions outside Africa and Australia also exhibit high runoff variability, but for different reasons—primarily erratic precipitation regimes rather than deeply weathered soils—with notable examples in Southwest Asia, the Brazilian Nordeste, and parts of the U.S. Great Plains.
Possible water reservoirs inside Earth
Geoscientific hypotheses propose that appreciable amounts of hydrogen and oxygen are sequestered within the solid Earth—in the crust, mantle and perhaps the core—and that these deep stores participate in a “whole‑Earth” water cycle that transfers volatiles between surface oceans and interior reservoirs over geological timescales. This concept implies dynamic, long‑term exchange of water constituents between surface and deep reservoirs rather than confinement of water to surficial liquid phases alone.
The volume of interior water remains highly uncertain and is actively debated; no single, universally accepted measurement exists. Some published estimates place the interior inventory at roughly 1.5 to 11 times the mass of the oceans, with that material residing hundreds of kilometres beneath the surface. Crucially, much of this proposed reservoir would not exist as free liquid but as non‑free water—for example structurally or chemically bound within minerals—so its physical state and mobility differ fundamentally from oceanic water.
If these larger estimates are correct, deep interior water would constitute a major, previously underappreciated component of the global hydrologic and geochemical system, with important implications for Earth’s total water budget and for models of long‑term surface–interior exchange processes.
Water in Earth’s mantle
Read more Government Exam Guru
In the mantle transition zone (roughly 520–660 km depth) high‑pressure silicate phases—especially wadsleyite and ringwoodite—can incorporate several weight percent of water as structurally bound hydroxyl (OH) within their crystal lattices. Ringwoodite in particular is a dominant phase in the mid‑transition zone and therefore constitutes a major deep reservoir for hydrogen in the form of OH defects rather than free molecular H2O. These nominally anhydrous minerals, when enriched in hydroxyl, measurably modify rock properties: they reduce effective viscosity, facilitate partial melting, lubricate plate interfaces, and lower seismic velocities.
Quantitative estimates of mantle water remain uncertain but large: some studies suggest that the lower mantle and transition zone together may store amounts of water comparable to, or exceeding, the entire surface inventory of the oceans—possibly several times that mass and in some scenarios amounting to tens of ocean equivalents. Direct mineralogical and geophysical evidence supports a substantive deep water reservoir. In 2014 a hydrous ringwoodite inclusion trapped in a Juína (Brazil) diamond provided mineral‑scale proof of OH‑bearing ringwoodite at transition‑zone pressures; super‑deep diamonds have also yielded ice‑VII and other high‑pressure water phases, demonstrating that distinct physical forms of water can exist at extreme depths. Complementary seismic observations, such as signals interpreted as dehydration‑melting near the top of the lower mantle beneath parts of the United States, indicate active mobilization or release of this bound water at the transition‑zone/lower‑mantle boundary.
Taken together, mineralogical, petrological and geophysical lines of evidence imply that hydrogen stored as hydroxyl in mantle minerals is a significant component of Earth’s volatile budget, with important consequences for mantle rheology, melting behavior and the long‑term cycling of water between interior and surface reservoirs.