Soil water content

Owned by Maiwenn Herpe
Last updated: Jun 24, 2021 by Frederika Mourot
5 min read

Satellite(s)

Landsat, MODIS, AVHRR, Sentinel-1, Sentinel-2, SMAP, SMOS.

Monitoring element

Land surface reflectance, SAR backscatter.

Satellite(s)

Landsat, MODIS, AVHRR, Sentinel-1, Sentinel-2, SMAP, SMOS.

Monitoring element

Land surface reflectance, SAR backscatter.

Description technique

Several remote sensing techniques exist for estimate soil water content. Optical soil moisture estimation takes advantage of the empirical prediction that drier soils will have higher reflectance within the visible spectrum that saturated soils. Common vegetation indices such as Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) have been considered for determining soil water content, giving a rapid but relative measure across large scales via open access sensors such Sentinel-2 and Landsat, which have the required visible, NIR and SWIR bands (Gu et al, 2008).

Thermal IR methods utilize the 'triangle method', which compares surface temperature and NDVI derived fractional vegetation coverage to predict soil moisture content (Babaeian et al, 2019). This model has been successfully implemented with sensors such as Landsat 8, as well as coarse resolution sensors such as MODIS and AVHRR, which offer the thermal bands required as inputs. Recently improved versions of this approach have been developed, such as OPTRAM, which can predict soil moisture with purely optical inputs by exploiting the relationship between soil water content and shortwave infrared reflectance (Babaeian et al, 2018).

Spaceborne SAR has also been leveraged for soil water content estimation, taking advantage of the strong contrast in dielectric response between dry soil and water (Gao et al, 2017). Active L, C and X bands from sensors such as RADARSAT, ERS and TerrraSAR-X have been applied to soil water content estimation. One approach is the change detection method, which takes advantage of multiple observations to reduce the impact of surface roughness. This approach compares the backscatter at any given point in a time series relative to the driest signal, with an associated NDVI value derived from coincident optical imagery of a similar observation date (Gao et al, 2017). By estimating the soil water content maximum and minimum values, this difference can be used to estimate soil water content per pixel.

Passive RADAR sensors such as SMOS and SMAP produce soil water content estimates using brightness and temperature measurements but at a very low spatial resolution (>35km), trading the higher resolution of active sensors for the greater water content sensitivity of passive sensors (Barrett and Petropoulos, 2013).

Accuracy / Resolution

Variable spatial and temporal resolution according to sensors.

Case study

Babaeian et al (2018) compared soil moisture estimates from the OPTRAM to Cosmic ray neutron (CRN) ground data across five sites. Results showed that the optical satellite derived soil water content estimates compared reasonably well, with a RMSE of 0.050 cm3/cm3, while also reflecting temporal moisture content dynamics well.

Gao et al (2017) demonstrated the SAR change detection method, combining with Sentinel-2 optical data to produce soil water content estimates at an 100m averaged resolution in Catalunya, Spain. The results compare well to ground measurements, though the authors note greater uncertainty is present in heavily vegetated sites. The averaging step was used to reduce the impact of surface roughness on the results.

SMAP has tested accuracy of ±0.04 m3/m3 in top 2-5 cm for vegetation water content < 5 kg/m2.

Benefits

  • Through the use of recently deployed sensors such as Sentinel-1, soil water content estimation can be applied at a much broader scale, and at a higher spatial and temporal resolution than can be achieved using field measurement or ground stations.

  • Soil water content measurements can be estimated within environments which would otherwise be inaccessible.

Limitations

  • Spaceborne multispectral sensors have issues penetrating vegetation land covers, with the spectral reflectance of soil required to estimate water content. A large number of factors beyond soil water content can also impact the spectral characteristics of the soil (Babaeian et al, 2019).

  • Both optical and microwave based measurement of soil water content is limited by the fact that generally only the surface layers of soil, down to a few cm, are able to be measured (Gao et al, 2017, Babaeian et al, 2019).

  • SAR derived measurements of soil water content are highly sensitive to surface roughness (Babaeian et al, 2018).

  • Application of vegetation indices to soil water content monitoring has shown that while there are significant correlations exist, such an approach is highly dependent on the homogeneity of the overlying vegetation cover (Gu et al, 2008).

  • The SAR change detection approach described requires measurements of driest and wettest soil moisture values, though these can be estimated, this obviously impacts the validity of the results.

Applicability for Northland

Yes, possibly.

SAR approaches have proven to be successful at determining relative soil moisture changes, and have the benefit of not being impacted by cloudiness, allowing year round monitoring. Calibration of a long term time series against in-situ data would be required to provide absolute estimates of soil water content.

Optical data will be limited in coverage and temporal granularity by the persistent cloud cover in the region, particularly during the winter months. Mature cloud-masking techniques are directly available for open access multispectral data (e.g. Landsat and Sentinel-2). When using commercial data, care must be taken to ensure that there is sufficiently cloud free imagery available, as cloud masking is not as mature and ordering a large volume of imagery to ensure complete cloud free coverage between multiple observations can become cost prohibitive.

Publication references

Xing, C., Chen, N., Zhang, X. and Gong, J., 2017. A machine learning based reconstruction method for satellite remote sensing of soil moisture images with in situ observations. Remote Sensing, 9(5), p.484.
https://www.mdpi.com/2072-4292/9/5/484

Babaeian, E., Sadeghi, M., Jones, S.B., Montzka, C., Vereecken, H. and Tuller, M., 2019. Ground, proximal, and satellite remote sensing of soil moisture. Reviews of Geophysics, 57(2), pp.530-616.
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018RG000618

Vicente-Serrano, S.M., Pons-Fernández, X. and Cuadrat-Prats, J.M., 2004. Mapping soil moisture in the central Ebro river valley (northeast Spain) with Landsat and NOAA satellite imagery: a comparison with meteorological data. International Journal of Remote Sensing, 25(20), pp.4325-4350.
https://www.tandfonline.com/doi/abs/10.1080/01431160410001712990

Babaeian, E., Sadeghi, M., Franz, T.E., Jones, S. and Tuller, M., 2018. Mapping soil moisture with the OPtical TRApezoid Model (OPTRAM) based on long-term MODIS observations. Remote sensing of environment, 211, pp.425-440.
https://www.sciencedirect.com/science/article/pii/S003442571830186X?via%3Dihub

Gao, Q., Zribi, M., Escorihuela, M.J. and Baghdadi, N., 2017. Synergetic use of Sentinel-1 and Sentinel-2 data for soil moisture mapping at 100 m resolution. Sensors, 17(9), p.1966.
https://www.mdpi.com/1424-8220/17/9/1966

Gu, Y., Hunt, E., Wardlow, B., Basara, J.B., Brown, J.F. and Verdin, J.P., 2008. Evaluation of MODIS NDVI and NDWI for vegetation drought monitoring using Oklahoma Mesonet soil moisture data. Geophysical Research Letters, 35(22).
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2008GL035772

Barrett, B. and Petropoulos, G.P., 2013. Satellite remote sensing of surface soil moisture. Remote sensing of energy fluxes and soil moisture content, 85-120.
http://eprints.gla.ac.uk/109110/

Hornacek, M., Wagner, W., Sabel, D., Truong, H.L., Snoeij, P., Hahmann, T., Diedrich, E. and Doubková, M., 2012. Potential for high resolution systematic global surface soil moisture retrieval via change detection using Sentinel-1. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 5(4), pp.1303-1311.

https://ieeexplore.ieee.org/document/6182719