Copy for LLM[View as Markdown](https://docs.planet.com/data/planetary-variables/soil-water-content/techspec/)

# Technical Specification

warning

Sales for Planet’s 20m and 100m Soil Water Content products have been discontinued. Their technical specification is available under [Enhanced Resolution](https://docs.planet.com/data/planetary-variables/soil-water-content/enhanced-resolution).

## Overview

Soil water content (SWC) is an essential component of the hydrologic cycle. Also known as soil moisture, it refers to the water between the soil surface and the groundwater level in the unsaturated zone. This water is replenished by precipitation, irrigation, or capillary rise from deeper layers, and depleted by percolation, soil evaporation, and plant transpiration.

The Planet Soil Water Content product is based on the microwave radiation naturally emitted by the Earth. This passive microwave radiation is related to the water content in the soil and provides a direct measurement of water in the top layer of the soil. It is not sensitive to cloud cover and, therefore, allows for consistent, frequent, and global coverage.

The Planet patented algorithm enhances the spatial resolution of the passive microwave observations and the physical-based Land Parameter Retrieval Model retrieves the soil water content.

Planet provides SWC data at 1000 meter resolution. With an archive spanning over 20 years, the data is consistent across time and space, boasting a strong correlation with ground data and accurately reflecting weather patterns. It is a cost-effective and low-effort alternative for measuring soil water content with physical sensors, making it suitable for many use cases like agriculture, water management, and insurance.

![](/data/planetary-variables/soil-water-content/swc_time_series.webp)

*Figure 1: Measurements of SWC near Ahlen, in North Rhine-Westphalia, Germany. The measurements of SWC of a single point are plotted over 5 years with the baseline climatology, showing how soil water content levels compare to the expected average for the area.*

## Product Specifications

*Table 1: SWC 1000 m product specification*

| Data Resource         | SWC 1000 m L band                                                                                                                    | SWC 1000 m C band                                                                                       | SWC 1000 m X band                                                                                   |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------- |
| Source ID             | SWC-SMAP-L\_V5.0\_1000                                                                                                               | SWC-AMSR2-C\_V5.0\_1000<br />(2012 - present)<br /><br />SWC-AMSRE-C\_V5.0\_1000<br />(2002 - 2011)     | SWC-AMSR2-X\_V5.0\_1000<br />(2012 - present)<br /><br />SWC-AMSRE-X\_V5.0\_1000<br />(2002 - 2011) |
| Version               | 5.0                                                                                                                                  | 5.0                                                                                                     | 5.0                                                                                                 |
| Unit                  | m³/m³                                                                                                                                | m³/m³                                                                                                   | m³/m³                                                                                               |
| Sensing Depth         | \~5 cm                                                                                                                               | \~2 cm                                                                                                  | \~1 cm                                                                                              |
| Pixel Size            | 1000x1000 m                                                                                                                          | 1000x1000 m                                                                                             | 1000x1000 m                                                                                         |
| Temporal Resolution   | 0° latitude: 137 to 183 observations per year<br /><br />40° latitude: 183 to 228 observations per year                              | 0° latitude: 205 to 228 observations per year<br /><br />40° latitude: 274 to 292 observations per year | 0° latitude: 205 to 228 obs/year<br /><br />40° latitude: 274 to 292 obs/year                       |
| Overpass Time         | 06:00 local solar time                                                                                                               | 01:30 local solar time                                                                                  | 01:30 local solar time                                                                              |
| Geographical Coverage | Global                                                                                                                               | Global                                                                                                  | Global                                                                                              |
| Data Availability     | 2015-04-01 - Present<br /><br />Data Gaps:<br />2019-06-19 to 2019-07-24<br />2022-08-05 to 2022-09-24<br />2022-11-16 to 2022-11-18 | 2002-06-15 - Present<br /><br />Data gap:<br />2011-10-04 to 2012-07-25                                 | 2002-06-15 - Present<br /><br />Data gap:<br />2011-10-04 to 2012-07-25                             |
| Satellites Used       | AMSR-2, SMAP                                                                                                                         | AMSR-E, AMSR-2                                                                                          | AMSR-E, AMSR-2                                                                                      |
| NRT latency (p90)     | 72 hours                                                                                                                             | 24 hours                                                                                                | 24 hours                                                                                            |
| Archive latency       | Within 30 days after creating a subscription                                                                                         | Within 30 days after creating a subscription                                                            | Within 30 days after creating a subscription                                                        |

The sensing depth described in the table is an estimation of the average condition. Several studies (For example, [Schmugge et al., 1986](https://ieeexplore.ieee.org/abstract/document/4072415); [Wang et al., 1987](https://ieeexplore.ieee.org/document/4072692); [Owe et al., 1998](https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/98WR01469)) have shown that microwave-based soil water content penetration depth depends on the soil water content conditions and frequency. At 1.41 GHz the penetration depth varies from approximately 10 cm to 1 m for soil conditions ranging from saturated to dry, whereas at 10.7 GHz the penetration depth varies from less than a few mm to a little over 2 cm for similar conditions. Thus, the drier the soil, the deeper the sampling depth.

## Asset Properties

Each SWC observation is delivered as a SWC data asset (`*swc.tif`) with metadata provided in a separate quality flag asset (`*swc-qf.tif`).

*Table 2: Asset properties of SWC data resources*

| Asset Name | Band Name | Unit     | Type   | Typical Range | No Data Value | Scale | Format  |
| ---------- | --------- | -------- | ------ | ------------- | ------------- | ----- | ------- |
| swc        | Band 1    | m3/m3    | UINT16 | 0 - 1         | 65535         | 0.001 | GeoTIFF |
| swc        | Band 2    | m3/m3    | UINT16 | 0 - 1         | 65535         | 0.001 | GeoTIFF |
| swc-qf     | Band 1    | unitless | UINT16 | NA            | 0             | 1     | GeoTIFF |

## Methodology

### Passive Microwaves and Downscaling Method

The Planet Soil Water Content data is derived from satellite sensors that measure passive microwave radiation. Passive microwave technology detects natural microwave emissions from the Earth's surface, which are sensitive to soil water content.

Using a patented algorithm ([US10643098](https://patentimages.storage.googleapis.com/1a/6f/1c/8cf51399ee7eaa/US10643098.pdf), [EP3469516B1](https://patentimages.storage.googleapis.com/f1/6c/0b/55cad50731d3e0/EP3469516B1.pdf)), we combine and downscale observations from multiple satellite constellation to a spatial resolution of 1000 m. We create footprints based on the footprint center, microwave frequency, incidence angle, azimuth angle, and footprint size, and then disaggregate them into equal interval ellipses using an internal Gaussian distribution. Within these ellipse-shaped footprints, the center contributes more to the observed values than the edges. By using fixed values for water bodies and filtering out influences such as radio frequency interference (RFI), we ensure accurate brightness temperatures.

### Physical-Based Land Parameter Retrieval Model

The downscaled high resolution microwave observations are converted to soil water content using the physical-based Land Parameter Retrieval Model (LPRM) ([Owe et al. 2001](https://ieeexplore.ieee.org/abstract/document/942542); [Owe et al. 2008](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2007JF000769); [De Jeu et al. 2014](https://www.sciencedirect.com/science/article/abs/pii/S0022169414001139), [Van der Schalie et al., 2018](https://www.mdpi.com/2072-4292/10/1/107), [Van der Schalie et al., 2018](https://www.mdpi.com/2072-4292/13/13/2480)).

The basis for the retrieval relies on the fact that thermal radiation in the microwave region is emitted by all natural surfaces as a function of temperature. The propagation of microwaves through the soil is then dependent on the soil dielectric constant, which has an almost linear relationship with soil water content. Microwave technology is the only remote sensing method that measures a direct response to the absolute amount of water in the soil, with a sensing depth ranging from millimeters for 37 GHz to tens of cm for 1.4 GHz. Vegetation also contributes to the microwave signal by absorbing, reflecting, and emitting radiation.

LPRM solves the radiative transfer equation at different frequencies and polarizations to retrieve soil water content and is currently the baseline algorithm for the ESA Climate Change Initiative (CCI) soil moisture products ([Scanlon et al., 2021](https://esa-soilmoisture-cci.org/sites/default/files/documents/public/CCI%20SM%20v06.1%20documentation/ESA_CCI_SM_RD_D2.1_v2_ATBD_v06.1_issue_1.1.pdf)).

### Input Data

*Table 3: List of inputs for Soil Water Content production*

| Product                        | Description                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        |
| ------------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Brightness Temperature L band  | Soil Moisture Active Passive (SMAP) Level-1B Radiometer Half-Orbit Time-Ordered L band Brightness Temperatures, Version 5 (downloaded from [NSIDC](https://nsidc.org/data/spl1btb/) in HDF5 format). This Level-1B product provides calibrated estimates of time-ordered geolocated brightness temperatures at 1.41 GHz with a footprint size of 39x47 km. SMAP L band brightness temperatures are referenced to the Earth's surface with undesired and erroneous radiometric sources removed. Data has been available since April 2015 with a latency of 12 hours. Detailed information is available [here](https://nsidc.org/sites/default/files/smap_atbd_radiometerl1b_tb_rev-b_20150401.pdf). |
| Brightness Temperature X band  | Advanced Microwave Scanning Radiometer for EOS (AMSR-E) and Advanced Microwave Scanning Radiometer 2 (AMSR-2) Level-1B Radiometer X band Brightness Temperatures (downloaded from [JAXA G-portal](https://gportal.jaxa.jp/gpr/?lang=en) in HDF5 format). This Level-1B product provides calibrated estimates of geolocated brightness temperatures at 10.7 GHz with a footprint size of 24x42 km. Data has been available from July 2002 to October 2011 (AMSR-E) and from June 2012 (AMSR-2) up to now with a latency of 12 hours. Detailed information is available [here](https://gportal.jaxa.jp/gpr/assets/mng_upload/GCOM-W/AMSR2_Level1_Product_Format_EN.pdf)                              |
| Brightness Temperature C band  | AMSR-E and AMSR-2 Level-1B Radiometer C band Brightness Temperatures (downloaded from [JAXA G-portal](https://gportal.jaxa.jp/gpr/?lang=en) in HDF5 format). This Level-1B product provides calibrated estimates of geolocated brightness temperatures at 6.9 GHz with a footprint size of 35x62 km. Data has been available from July 2002 to October 2011 (AMSR-E) and from June 2012 (AMSR-2) up to now with a latency of 12 hours. Detailed information is available [here](https://gportal.jaxa.jp/gpr/assets/mng_upload/GCOM-W/AMSR2_Level1_Product_Format_EN.pdf).                                                                                                                          |
| Brightness Temperature Ka band | AMSR-E and AMSR-2 Level-1B Radiometer Ka band Brightness Temperatures (downloaded from [JAXA G-portal](https://gportal.jaxa.jp/gpr/?lang=en) in HDF5 format). This Level-1B product provides calibrated estimates of geolocated brightness temperatures at 36.5 GHz with a footprint size of 7x12 km. Data has been available from July 2002 to October 2011 (AMSR-E) and from June 2012 (AMSR-2) up to now with a latency of 12 hours. Detailed information is available [here](https://gportal.jaxa.jp/gpr/assets/mng_upload/GCOM-W/AMSR2_Level1_Product_Format_EN.pdf).                                                                                                                         |
| Brightness Temperature W band  | AMSR-E and AMSR-2 Level-1B Radiometer W band Brightness Temperatures (downloaded from JAXA G-portal in HDF5 format). This Level-1B product provides calibrated estimates of geolocated brightness temperatures at 89 GHz with a footprint size of 3x5 km. Data has been available from July 2002 to October 2011 (AMSR-E) and from June 2012 (AMSR-2) up to now with a latency of 12 hours. Detailed information is available [here](https://gportal.jaxa.jp/gpr/assets/mng_upload/GCOM-W/AMSR2_Level1_Product_Format_EN.pdf).                                                                                                                                                                     |
| Digital Elevation Model        | Digital elevation model (DEM) static map resampled at 100 m based on the Copernicus DEM GLO-90 product covering the full global landmass of the time frame of data acquisition (2011-2015). Detailed information is available in the link.                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
| Land Cover Map                 | Custom global land classification including permanent water bodies based on the Copernicus Global Surface Water Bodies product from PROBA-V. Detailed information is available [here](https://land.copernicus.eu/global/)                                                                                                                                                                                                                                                                                                                                                                                                                                                                          |
| Soil Map                       | Soil property maps resampled at 100 m based on the SoilGrids product. SoilGrids was funded by the core funding of ISRIC with additional support from the EUH2020 CIRCASA project. Detailed information is available [here](https://www.isric.org/explore/soilgrids),                                                                                                                                                                                                                                                                                                                                                                                                                               |

## Data Quality

### Uncertainty

The accuracy of soil water content retrievals from passive microwave observations and in particular the performance of the Land Parameter Retrieval Model has been described widely in the scientific literature: it has been compared to ground stations, hydrological models, and other satellite-derived products (for example, [Van der Schalie et al., 2021](https://www.mdpi.com/2072-4292/13/13/2480); [Gruber et al., 2020](https://www.sciencedirect.com/science/article/pii/S0034425720301760); [Gevaert et al., 2018](https://hess.copernicus.org/articles/22/4605/2018/)).

The accuracy of the retrievals depends on the microwave frequency. L-band soil water content has the highest accuracy of approximately 0.04 m³/m³, which is comparable to the accuracy of ground sensors (for example, see [Ganjegunte et al., 2012](https://link.springer.com/article/10.1007/s13201-012-0032-7)). The accuracy of the C and X band based products is approximately 0.05 m³/m³. In addition, the accuracy is a direct function of the vegetation cover as demonstrated by [Van der Schalie et al., 2018](https://www.mdpi.com/2072-4292/10/1/107).

The X and C band based retrievals are most sensitive to the density of the vegetation cover and provide the most reliable values with short vegetation. L band retrievals perform well over the majority of land covers, with the exception of dense tropical rainforest.

### Validation

We have evaluated our SWC products by comparing them to carefully picked ground data stations across various land cover types. The Pearson correlation coefficient varied between 0.53 and 0.92, indicating strong relations with ground truth data. Figure 2 shows a comparison between SWC data and ground data.

You can read more about the results in [this white paper](https://planet.widen.net/s/drrxtldpwg).

![](/data/planetary-variables/soil-water-content/swc_validation.webp)

*Figure 2: Time series of the average of 9 stations of the RAAM network, the Netherlands.*

### Metadata

Quality flag assets (`*swc-qf.tif`) provide metadata for each pixel using bitwise flags. Critical flags indicate unreliable data, with corresponding pixels in band 1 of the SWC asset (`*swc.tif`) set to the no data value. The replaced SWC value can be found in band 2 of the SWC asset. Non-critical flags indicate that the data can be used with caution, taking into account the flag description.

Critical and non-critical flags are described in the tables below. For more information on how to access the quality flag asset, check out [subscribing to planetary variables](https://docs.planet.com/develop/apis/subscriptions/sources.md).

*Table 4: Non-critical flags*

| Bit | Flag layer                    | Description                                                                                                                                                                                                                                 |
| --- | ----------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| 1   | Dense vegetation              | The retrieved soil water content is less reliable over dense vegetation cover.                                                                                                                                                              |
| 2   | Low soil water content        | The retrieved soil water content is lower than the estimated wilting point.                                                                                                                                                                 |
| 3   | High soil water content       | The retrieved soil water content is higher than the estimated porosity.                                                                                                                                                                     |
| 4   | Possible severe precipitation | Part of the footprints touch an area flagged as severe precipitation.                                                                                                                                                                       |
| 5   | Possible RFI                  | Footprints are contaminated for less than 25% with Radio Frequency Interference (RFI). RFI occurs when human-made transmitters emit in the same frequencies and thus disrupt the radiometer measurements of the natural microwave emission. |
| 7   | Possible frozen soil          | The soil may be frozen. These are pixels with a soil temperature between -10°C and 0°C.                                                                                                                                                     |

*Table 5: Critical flags*

| Bit | Flag layer                                | Description                                                                                                                                                                                                                                 |
| --- | ----------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| 6   | Statistical outlier                       | The underlying footprint is considered a statistical outlier.                                                                                                                                                                               |
| 8   | Frozen soil                               | The soil is frozen. Pixels with a soil temperature below -10°C. A conservative value of -10°C was chosen to avoid masking valid data.                                                                                                       |
| 9   | Severe precipitation                      | Severe precipitation is detected.                                                                                                                                                                                                           |
| 10  | Vegetation too dense                      | The vegetation cover is too dense for the algorithm to reliably retrieve soil water content.                                                                                                                                                |
| 11  | No overpass                               | The satellite did not pass over.                                                                                                                                                                                                            |
| 12  | RFI                                       | Footprints are contaminated for more than 25% with Radio Frequency Interference (RFI). RFI occurs when human-made transmitters emit in the same frequencies and thus disrupt the radiometer measurements of the natural microwave emission. |
| 13  | Instrumental flaws                        | Unrealistic values due to instrumental flaws. If the brightness temperature at 36.5 GHz V produces values either over 400K or under 0.9 \* the water temperature, the data is considered as unrealistic and is removed.                     |
| 14  | Out of valid range                        | Soil water content values are outside the valid range, meaning under 0 m3/m3 or above 1 m3/m3.                                                                                                                                              |
| 15  | Open water                                | Soil water content is not defined over water.                                                                                                                                                                                               |
| 16  | Brightness temperature residuals too high | The LPRM model could not find a soil water content value that is consistent with all provided input data.                                                                                                                                   |

![](/data/planetary-variables/soil-water-content/swc_quality_flags.webp)

*Figure 3: An example of two quality flags for a region around Nantes, France. The image shows critical flags 'open water' and 'no overpass' overlaid on SWC.*

The following Python script converts a quality flag pixel value into a list of corresponding quality flags. Note that one pixel may have multiple flags applied to it.

```
# swc_quality_flags.py



import argparse



SWC_QUALITY_FLAGS = {

    1: "Dense vegetation",

    2: "Low soil water content",

    3: "High soil water content",

    4: "Possible severe precipitation",

    5: "Possible RFI",

    6: "Statistical outlier (critical flag)",

    7: "Possible frozen soil",

    8: "Frozen Soil (critical flag)",

    9: "Severe precipitation (critical flag)",

    10: "Vegetation too dense (critical flag)",

    11: "No overpass (critical flag)",

    12: "RFI (critical flag)",

    13: "Instrumental flaws (critical flag)",

    14: "Out of valid range (critical flag)",

    15: "Open water (critical flag)",

    16: "Brightness temperature residuals too high (critical flag)",

}





def convert_to_quality_flags(decimal_value: int) -> list[str]:

    binary_string = format(decimal_value, "016b")

    reversed_binary_string = binary_string[::-1]

    return [

        f"{i}. {SWC_QUALITY_FLAGS.get(i, 'Unused flag')}"

        for i, bit in enumerate(reversed_binary_string, start=1) if bit == "1"

    ]





if __name__ == "__main__":

    parser = argparse.ArgumentParser(description="Convert a decimal value to quality flags.")

    parser.add_argument("decimal_value", type=int, help="The decimal value to convert.")

    args = parser.parse_args()



    flags = convert_to_quality_flags(args.decimal_value)

    print(*flags, sep="\n")
```

that you can call as follows, using value `32770` as example:

```
> python swc_quality_flags.py 32770

2. Low soil water content

16. Brightness temperature residuals too high (critical flag)
```

## Other Limitations

The following are some key limitations that we have not flagged:

* SWC data should be considered unreliable over sloped terrain that is over 20 degrees. The microwave signal can be distorted by the angle of the slope, causing changes in the observed brightness temperature. This distortion leads to inaccuracies in the soil water content estimates because the algorithms assume a flat surface. Data usually shows up unrealistically dry.
* SWC data should be considered unreliable over urban areas. Human-made surfaces like concrete and asphalt have different microwave emissivity properties than natural soil. These surfaces can cause reflections and scattering of the microwave signal, leading to incorrect soil water content retrieval. Moreover, because urban areas have such a wide range of land cover types (asphalt, concrete, roofs, parks, gardens, etc) the average SWC value is not representative of any specific land cover type.
* Dynamic water bodies may cause unrealistic soil water content estimates. Water bodies influence the microwave observations and without proper mitigation they could result in unrealistic retrievals. Data around dynamic waterbodies can show up drier or wetter.
* False cloud/shadow detections may occur if surface conditions change very rapidly, during prolonged cloudiness, or over AOIs with significant terrain and shadowing. Significant effort has gone into developing automated techniques to differentiate between actual change and atmospheric contamination, but there may still be false detections.
* The SWC product integrates passive microwave data streams from multiple satellite missions. In rare instances where source data is completely unavailable, Planet cannot calculate SWC for that specific date. Once the 30-day reanalysis window closes, any such missing date in the subscription will be designated with a READY\_NO\_DATA status.

## Frequently Asked Questions

### There are three different bands for soil water content. What is a 'band' and which one should I use?

We use passive microwave observations to measure soil water content. A microwave signal is a light signal with a frequency that is much lower than the frequency of visible light. The lower the frequency of the passive microwaves, the more they tell us about what happens below the ground. A “band” refers to the frequency band we use to measure SWC. The lowest frequency band we deliver is the L-band (1.4 GHz), which is measured by the SMAP satellite. This measurement shows the water content in the top 5 cm of the soil. Other frequencies we deliver are the C-band (6.9 GHz), which shows the water content in the top 2 cm, and X-band (10.7 GHz), showing the top cm of the soil. The X-band and C-band measurements come from the AMSR-2 satellite.

For most use cases, L-band measurements are the best choice:

* They are representative of a deeper layer than the other bands.
* Vegetation on top of the soil has a much lower impact on the signal than for other bands.
* L-band observations are more sensitive to soil water content changes than the other bands: per unit change in soil water content, the microwave soil emission changes are stronger for L-band than for higher frequencies like C-band. This higher sensitivity leads to more accurate and precise SWC observations.

There are however some situations where the other bands are a better choice:

* Sometimes, radio signals emitted from the Earth interfere with the soil water content signals. We call this effect “radio frequency interference” (RFI). We filter for RFI, but sometimes the filter does not filter out everything, leading to strange SWC values. When this happens in the L-band for a specific location, use X-band or C-band.
* AMSR-2 provides the observations for X-band and C-band SWC. AMSR-2 data is available from 2012 onwards, and can be combined with data from its predecessor (AMSR-E) to compute long SWC time series, all the way back to 2002. When a customer wants to have a long time series, C-band and X-band data can go back to 2002. Our L-band product can provide SWC data from 2015 to present.

### Is SWC a modeled product or an observation?

The brightness temperatures are direct observations by the satellite. These observations are used in a physical-based radiative transfer model to retrieve soil water content. However, considering the strong physical description of the radiative transfer model, scientists often refer to passive microwave soil water content as satellite-observed.