Soil moisture can be calculated in two fundamental ways: by weight (gravimetric method, comparing wet and dry soil mass) or by volume (volumetric method, expressing water as a fraction of a given soil volume). The gravimetric approach requires only an oven and a scale, while volumetric readings increasingly come from electronic sensors buried in the ground, neutron-based instruments, satellite imagery, or simple water-balance equations driven by weather data. Which method and formula you use depends on whether you are managing a backyard garden, scheduling irrigation on a commercial farm, or validating satellite data across a watershed.
The Gravimetric Method
This is the oldest and most straightforward technique. You collect a soil sample, weigh it, dry it in an oven at about 105 °C for 24 hours, and weigh it again. The difference is the mass of water that was in the sample. Gravimetric water content is simply the mass of that water divided by the mass of the oven-dried soil, usually expressed as a percentage or as grams of water per gram of dry soil.
The method is cheap, requires no calibration, and serves as the reference standard against which nearly every sensor and model is ultimately compared. Its downsides are that it destroys the sample (you cannot re-measure the same spot), it takes a full day to get results, and it tells you nothing about moisture at a particular moment in time if conditions are changing quickly. For research plots where you need a dependable ground truth, though, it remains the benchmark.
Converting Gravimetric to Volumetric Water Content
Most practical applications, from irrigation scheduling to hydrological modeling, need volumetric water content: the volume of water per volume of soil. Converting from gravimetric to volumetric requires knowing the soil’s bulk density, which is the mass of dry soil per unit volume. You multiply the gravimetric water content by the bulk density (and divide by the density of water, which is effectively 1 g/cm³) to get volumetric water content.
This sounds simple, but bulk density is not constant. It changes with depth, compaction, tillage, and even with the moisture level itself in soils that shrink and swell. In cracking clay soils, ignoring the volume change that happens as the soil dries can overestimate volumetric water content by as much as 24%.1Journal of Soil Science. Calculation of the field volumetric water content of cracking clay soils from measurements of gravimetric water content and bulk density In one study of a 1.5-meter-deep soil profile, using a single bulk density value for the whole profile led to an underestimation of total stored water by about 35 mm, roughly a 7% error.2Vadose Zone Journal. Methods for downhole soil water sensor calibration—Complications of bulk density and water content variations For soils that do not crack or swell much, a one-time bulk-density measurement at each depth is usually good enough. For expansive clays, you need to account for how the soil volume itself changes as it wets and dries.
Electronic Sensors That Read Moisture in Place
Rather than pulling samples out of the ground, most modern monitoring relies on sensors that stay buried and report readings continuously. They all work by measuring some property of the soil that changes with water content, then converting that measurement to a moisture value using a calibration equation.
Time Domain Reflectometry and Related Dielectric Methods
TDR sensors send an electromagnetic pulse along metal rods (called waveguides) buried in the soil and measure how fast the pulse travels. Water has a much higher dielectric constant than dry soil minerals or air, so wetter soil slows the pulse noticeably. A widely used empirical equation developed by Topp and colleagues in 1980 converts the measured dielectric constant directly to volumetric water content. It works well in many mineral soils without any site-specific calibration. In a comparison study across four soil textures, a dielectric mixing model provided the best overall fit to measured moisture, with an error of about 0.027 m³/m³, though in sandy loam the simpler Topp equation performed just as well.3Elsevier. Comparison of soil water content estimation equations using ground penetrating radar
Capacitance and FDR Sensors
Capacitance sensors operate on a related principle: they measure the dielectric properties of the surrounding soil using oscillating electric fields rather than a traveling pulse. They tend to be less expensive than TDR and come in many form factors, from small probes you push into a pot to long tubes that read at multiple depths. Their accuracy, however, is more sensitive to the specific soil they are installed in, and factory calibrations can be significantly off for soils with unusual textures or high organic matter.
A key vulnerability is air gaps. If the sensor does not make perfect contact with the surrounding soil, the reading drops because air has a very low dielectric constant. Testing has shown that a commercially available capacitance sensor significantly underestimated moisture when even small air gaps were simulated around it, while a newer spatial-frequency-domain sensor was almost unaffected by the same gaps.4PubMed Central. Low-Error Soil Moisture Sensor Employing Spatial Frequency Domain Transmissometry If you are installing capacitance probes, packing soil tightly around the sensor and avoiding loose backfill is essential for reliable readings.
Electrical Resistance Blocks
Gypsum blocks and similar resistance-based sensors take a different approach. Two electrodes are embedded in a porous block that is buried in the soil. As the surrounding soil wets or dries, the block equilibrates to the same moisture tension, and the electrical resistance between the electrodes changes. These sensors do not measure water content directly; they measure soil water potential, which tells you how tightly the soil is holding onto its water. A digital meter designed for gypsum blocks provides good resolution of matric water potential in the range of about −0.06 to −1.5 MPa, covering the range from fairly wet soil to near the permanent wilting point.5Soil Science Society of America Journal. An Economical, Digital Meter for Gypsum Soil Moisture Blocks They are inexpensive and well-suited for irrigation scheduling where you care more about whether roots can extract water than about the exact percentage of moisture present.
Tensiometers and the Water-Potential Approach
A tensiometer works by letting a porous ceramic tip come into equilibrium with the surrounding soil water. The suction that develops inside the tube is a direct reading of soil matric potential. Traditional tensiometers top out around −0.08 to −0.09 MPa, limiting them to relatively wet soils. Polymer tensiometers push this range much further, directly measuring matric potentials down to about −1.6 MPa, close to the theoretical wilting point.6Hydrology and Earth System Sciences. Polymer tensiometers with ceramic cones: direct observations of matric pressures in drying soils When combined with TDR readings at the same location, they allow you to build a soil water retention curve in the field, which maps the relationship between water content and water potential for your specific soil.
Soil water retention curves are critical for translating between the two languages of soil moisture: “how much water is there” (volumetric content) and “how available is that water to plants” (matric potential). The widely used van Genuchten model describes these curves mathematically, and researchers have extended it to account for changes in bulk density so the same curve framework can apply even as the soil compacts or loosens over time.7Water Resources Research. Approaches for Estimating Soil Water Retention Curves at Various Bulk Densities With the Extended Van Genuchten Model
Neutron Probes
The neutron probe has been a workhorse in soil moisture research for decades. A radioactive source lowers into an access tube emits fast neutrons, which collide with hydrogen atoms in the soil (primarily in water molecules) and slow down. A detector counts these slowed (“thermalized”) neutrons, and the count rate is proportional to volumetric water content after calibration.8Canadian Geotechnical Journal. Soil moisture monitoring at the field scale using neutron probe
The technique measures a sphere of soil roughly 15 to 30 cm in radius, which gives a more integrated reading than a small point sensor. It is also less affected by cracking in expansive soils than buried probes, because the measurement sphere is large enough to average across cracks.9Geotechnical Testing Journal. Water Content Measurement in Expansive Soils Using the Neutron Probe However, the readings are sensitive to the diameter of the access hole and how well the tube fits, so careful installation matters.10Soil Science Society of America Journal. Effect of Access Hole Properties on Soil Water Content Determination by Neutron Thermalization The main practical drawback is that neutron probes contain a radioactive source, which means licensing, safety training, and regulatory compliance. For this reason, they have gradually given way to electronic sensors for routine use, though they remain valuable for research and for validating other instruments.
Cosmic-Ray Neutron Sensing
A more recent innovation uses naturally occurring cosmic-ray neutrons rather than an artificial source. Cosmic rays from space continuously produce fast neutrons when they strike atoms in the atmosphere and soil surface. Like a neutron probe, the count of slow neutrons returning above ground is related to how much hydrogen (and therefore water) is in the top layer of soil. A single stationary detector can sense moisture over a footprint roughly 200 to 300 meters in diameter, bridging the gap between point sensors and coarse satellite products.11European Journal of Soil Science. Capturing Field‐Scale Soil Moisture Dynamics in Ireland Using Cosmic‐Ray Neutron Sensing
The trade-off is depth. Cosmic-ray neutron sensors read the top 10 to 30 cm of soil, depending on how wet it is (wetter soil means a shallower sensing depth). For applications that need root-zone moisture down to a meter or more, this shallow reading has to be extended using depth-scaling methods or paired with point sensors at deeper levels.12Hydrology and Earth System Sciences. Estimating field-scale root zone soil moisture using the cosmic-ray neutron probe Performance in complex terrain, such as alpine environments with varying biomass and steep slopes, still requires careful correction, though multi-year evaluations on the Tibetan Plateau and elsewhere are refining these techniques.13Vadose Zone Journal. Field‐scale soil moisture estimation in complex alpine terrain using cosmic‐ray neutron sensors
Water Balance Equations
You do not always need to measure soil moisture directly. If you can track the water going into and out of the soil, you can calculate what must be left. The basic water balance for irrigation scheduling tracks soil moisture depletion as the running sum of evapotranspiration minus effective rainfall minus irrigation plus any drainage out the bottom of the root zone.14Transactions of the ASAE. Estimating Soil Moisture Depletion from Climate, Crop and Soil Data When depletion exceeds a threshold you set for your crop and growth stage, it is time to irrigate. The amount of water to apply equals the current depletion divided by your irrigation system’s efficiency.
This approach has practical appeal because the inputs, rainfall, temperature, and crop coefficients, come from weather stations and published tables rather than buried hardware. A quasi-terrestrial water balance method driven by common weather data has shown favorable agreement with field-measured soil moisture storage at monthly and seasonal time scales.15Agricultural Water Management. Estimating soil moisture storage change using quasi-terrestrial water balance method Validation against ground stations across India found correlation coefficients of 0.46 to 0.60, which is decent for a method that uses no sensors in the ground at all.16MAUSAM. Validation of soil moisture derived from water balance method and satellite observation Satellite-based evapotranspiration products are increasingly being plugged into these equations. An evaluation of six satellite ET models for irrigated alfalfa in Arizona found that their ensemble estimate was a reliable input for calculating soil water depletion in an arid climate.17Agricultural Water Management. Evaluation and comparison of OpenET models for estimating soil water depletion of irrigated alfalfa in Arizona
Remote Sensing for Broad-Scale Estimates
Satellites and drones provide moisture estimates over large areas without touching the soil, using microwave or thermal-optical techniques. Microwave sensors, both active (radar) and passive (radiometers), respond to the dielectric properties of the surface soil and can see through clouds. Thermal and optical approaches combine land surface temperature with vegetation indices to infer how dry the surface is. For instance, the Temperature Vegetation Dryness Index combines surface temperature and vegetation greenness from satellite imagery to monitor drought conditions, and has been evaluated over multi-decade records for regional drought tracking.18PubMed Central. Temperature vegetation dryness index (TVDI) for drought monitoring in the Guangdong Province from 2000 to 2019
The resolution of satellite-based moisture estimates remains coarse compared to what a single sensor measures, typically tens of kilometers for global microwave products. Validating them against ground truth requires careful spatial sampling strategies. Research into efficient sampling schemes has shown that algorithm-driven selection of sample locations, guided by terrain and remote sensing data, can capture the spatial variability of a field more accurately than simple random sampling.19Geoderma. An efficient soil moisture sampling scheme for the improvement of remotely sensed soil moisture validation over an agricultural field
Salinity and Other Interference Problems
Dielectric-based sensors, both TDR and capacitance types, assume that the speed or frequency response of the electromagnetic signal is driven primarily by water content. Dissolved salts in soil water raise the electrical conductivity of the soil, which can interfere with the measurement and cause overestimation of moisture. Capacitance sensors in particular are highly susceptible to temperature, salinity, and soil type effects.20PubMed Central. Advances in Calibration Methods for FDR-Based Capacitive Soil Moisture Sensors
A comparative test of eight commercial sensors at increasing salinity levels found that all overestimated moisture to varying degrees as salinity rose. Most sensors performed reasonably well when the soil’s electrical conductivity stayed at or below 1.0 dS/m, but several experienced sudden distortions at 3.0 dS/m. After site-specific calibration, accuracy could be brought to within about ±0.02 cm³/cm³ as long as measurements stayed below each sensor’s distortion threshold.21PubMed Central. Performance of Soil Moisture Sensors at Different Salinity Levels: Comparative Analysis and Calibration If you are working in saline or brackish conditions, sensor choice and calibration are not optional niceties; they determine whether your readings are meaningful at all.
High organic matter creates a similar headache. Soils amended with biochar, for example, caused all tested sensors to overestimate water content when using factory calibrations. After a simple linear recalibration, accuracy jumped to an R² of roughly 0.99 across three different sensor types.22Frontiers. Calibration and performance evaluation of soil moisture sensors in biochar-amended soils using an IoT-Based sensor system The lesson is consistent: factory equations get you in the right neighborhood, but soil-specific calibration against gravimetric samples is what gets you to research-grade accuracy.
Sensor Placement and Root-Zone Integration
A single sensor at one depth gives you a point measurement, but crops draw water from a range of depths. To estimate total stored water across a 100-cm profile, research recommends placing sensors at approximately 10, 30, 60, and 90 cm.23Vadose Zone Journal. Optimizing soil moisture sensor installation depths and number of sensors for in situ monitoring networks Each reading then represents a slab of soil, and you multiply the volumetric water content at each depth by the thickness of the slab it represents, then sum the results to get total profile storage in millimeters. Fewer sensors mean you are interpolating more, which introduces error especially in layered soils where texture changes sharply with depth.
Machine learning approaches are beginning to fill gaps between sensor depths by fusing point readings with remote sensing data. One study estimated multi-layer moisture in 30-cm increments down to 180 cm by combining ground sensors, satellite imagery, and weather data through machine learning models.24Vadose Zone Journal. Multi‐layer root zone soil moisture estimation using field and remote sensing data fusion with machine learning in semi‐arid croplands These hybrid methods are promising for commercial agriculture, where installing sensors every few centimeters is impractical but knowing how deep the wetting front has reached matters for irrigation decisions.
How Many Measurement Points Does a Field Need
Soil moisture varies dramatically across even a single field because of small differences in texture, slope, compaction, and root density. A single probe tells you what is happening at that exact spot, not what the average field moisture is. So how many points do you actually need?
Studies using randomized sampling methods have found that the necessary sample size depends heavily on the size and variability of the area. In experimental plots ranging from 10 × 10 m up to 160 × 160 m, the required number of samples increased with plot size and with the spatial variability present at the time of measurement.25Journal of Hydrology. Estimating the necessary sampling size of surface soil moisture at different scales using a random combination method For mapping changing moisture patterns on turfgrass using machine learning, 25 to 40 sensor readings were sufficient on most dates, though the recommendation was to deploy 40 to 60 sensors to allow for signal drops and hardware failures.26Agronomy. Investigation of Sample Numbers Needed to Map Spatial Changes in Soil Moisture Using Random Forests and Z-Score Calibration for Precision Irrigation of Turfgrass If you are managing a large irrigated field and want to move beyond single-point monitoring, understanding this sampling question is just as important as choosing the right sensor technology.