## What type of satellites are used in your algorithms?¶

We use passive microwave satellites, which measure microwave signals that are naturally radiating from the Earth’s surface. This enables observations to be acquired during cloudy conditions, because of the physical properties of waves transmitted in this spectrum’s range. Examples of the satellites/sensors we use are:

## When is the data observed?¶

Currently, VanderSat uses the nighttime observations during the descending orbits. Each satellite has its own time of observing an area, which is roughly the same time every day. For the satellites we use, these overpasses are either around 01:30 solar time or 06:00 solar time. The data acquired at that time can be regarded as a snapshot of the soil conditions at that time. Therefore, the measurements we provide are representative for the time of overpass.

## What are the depths of the measurements?¶

The sensors we use measure the signal originating from the top layer of the soil. Typically, this is up to 10 cm deep, though the strongest contribution is received from the most upper layer. Commonly assumed is a depth of roughly 5 centimeter, but in reality the depth of this measurement varies slightly with moisture content. If the soil is drier, the sensor can see deeper into the soil.

## Can you measure anything deeper than the top layer of the surface?¶

Not directly. A direct measurement is only possible of the top layer of the soil. However, there are ways of translating this surface measurement into something representative for deeper layers as well. This always requires some sort of model, which is inherently limited to its assumptions. A very simple, yet powerful way to do is the Derived Root Zone Soil Moisture calculation that VanderSat delivers. Through this calculation, one can approximate the water content for the root zone up to 50 cm deep. See also Derived Root Zone Soil Moisture.

## What does the unit $$m^3 m^{-3}$$ represent?¶

The unit $$m^3 m^{-3}$$ of the VanderSat data products is volumetric soil moisture, which indicates the volume fraction of water in a volume of soil. A value of 0.4 is equivalent to 400 liters of water in one $$m^3$$ of soil. See also Soil Moisture.

## Where can I find how your soil moisture retrieval works?¶

VanderSat’s retrieval algorithm is based on the Land Parameter Retrieval Model (LPRM), which has been extensively described in the scientific literature (see Scholar Search). VanderSat has taken this well-tested method and uses its patented algorithm to go to a much higher spatial resolution. See also High resolution satellite soil moisture.

## What is your accuracy and precision?¶

The precision of the VanderSat soil moisture data is about 0.001 $$m^3/m^3$$. Accuracy is lower at approximately 0.03 $$m^3/m^3$$. This is similar to properly installed in-situ sensors (see e.g. this paper ). This said, there are no independent measurements at the scale we are observing (e.g. no gravimetric measurements at 100x100m) so the true accuracy remains largely unknown.

## Is it a modelled product or an observation?¶

The brightness temperatures are direct observations by the satellite. These are used in combination with a dielectric mixing model (LPRM, see Where can I find how your soil moisture retrieval works?) to retrieve soil moisture. As such, the final product is based on measurements but includes some modeling.

## What do you measure in cities?¶

The soil moisture values that we measure in urban areas are simply the measured dielectic constant values of the city’s surface converted into soil moisture. Although we do show soil moisture values over urban areas, these values don’t actually represent soil moisture. You can use the built-up area data flag to mask out cities and other urban areas if required. See How to retrieve the data flags?.