Dear All,

I want to point out that when you measure soil surface temperature with an infrared thermometer, you really get values from the surface. This is not the case with remote soil moisture measuring because there is always a depth involved.

I believe that "surface soil moisture" always means "near surface soil moisture", and perhaps that is the terminology that we should use.

Comparison of "near surface soil moisture" measured remotely and "in situ" is almost impossible because both sensors integrate differently (in depth and space) and in ways that therefore cannot be compared.

Best wishes, Kees

As water infills, substances more solubles go down and make a differentiation in the composition.

Dear Mateusz,

You asked an appealing question. The most soil moisture data from remote sensing (RS) are limited to the near-surface soil layer which considered as the above 5 or 10 cm of the soil profile. You are right that this method estimate the average moisture within this particular depth. Similarly, the direct measurements, for example excavation, also determine the average soil moisture within this 5 or 10 cm depending on the height of the measurement device (core sampler). Therefore, these two methods applied the same procedure to find the soil moisture.

Maybe you can find more at:

Mohanty et al (2001). Hydraulic parameter estimation by remotely-sensed top soil moisture observations with the particle filter

Dear Mateusz,

Surface soil moisture is changing not only with depth but also with time and space. Therefore, an answer on your question must presume the subject properly to your demands. In other words a very general frame model is needed for considering the surface intensity of soil moisture vs depth. In a simplest case you may assume a stable volumetric soil moisture state for excluding time variations. But soon you shall need estimating the representative surface soil moisture over some roi area. Thus we talk on two different meanings under surface sm and the stable value of volumetric sm under ground. The question - what answer is desired remains and probably you mean the need of a useful measure of the soil moisture state employing the volumetric soil moisture and distinguish it from the surface soil moisture. In fact your question is for a model and a related well defined space of soil moisture states and its norm. If you mean the static case, then things are relatively simple upon using the saturation soil moisture measures [0-100%] instead of sm in volumetric measures also spread in the range [0-100%] but meaning different measures. A couple between is available by means of hydrological and pedological parameters of soil. Soil science is more involved, but still there is a question about the uncertainty of available RS data. Please look how others managed about that trouble in the following articles. The model is not only a useful pattern ready for employing, there are many possible, but is also a way of defining our own task and subject and locating them properly in the domains of time and space.

http://www.researchgate.net/publication/236949566 C. Montzka, H. Moradkhani, L. Weihermüller, H-Jan Hendricks Franssen, M. Canty, H. Vereecken Hydraulic Parameter Estimation by Remotely- sensed top Soil Moisture Observations with the Particle Filter in JOURNAL OF HYDROLOGY · MARCH 2011

www.hydrol-earth-syst-sci.net/16/3499/2012/ V. Maggioni, E. N. Anagnostou, and R. H. Reichle, The impact of model and rainfall forcing errors on characterizing soil moisture uncertainty in land surface modeling, Hydrol. Earth Syst. Sci., 16, 3499–3515, 2012, doi:10.5194/hess-16-3499-2012

regards

Wojciech Marczewski

Article Hydraulic Parameter Estimation by Remotely-sensed top Soil M...

Dear All,

I want to point out that when you measure soil surface temperature with an infrared thermometer, you really get values from the surface. This is not the case with remote soil moisture measuring because there is always a depth involved.

I believe that "surface soil moisture" always means "near surface soil moisture", and perhaps that is the terminology that we should use.

Comparison of "near surface soil moisture" measured remotely and "in situ" is almost impossible because both sensors integrate differently (in depth and space) and in ways that therefore cannot be compared.

Best wishes, Kees

I would also add that it will depend on the time history of the recharge to the soil, whether there is a precipitation or evaporation cycle and for how long.  

Dear we have publish some articles shown soil moisture dynamics at surface soil (using TDR, soil thermometer, electronic tensiometers and lysimeters) as affected by different tillage and establishment method options available at the farmer level and we have observed that mulched happy seeder plots plots dried more than unmulched plots during intervening periods.

I think you are correct. In addition, not only the type of soil is important, but also the climate of the region. In our case (tropical country), we have a coastal humidity and temperature about 30ºC. The soils are most part humid since their surface to any depth, in this case, depending on the texture of the soil. By other side, in semi-arid portion of our region, the surface is hot at maximum sun radiation (about 14 PM) and the lower portion is more cold. Then, the inner portion will heat during the day, and at night the surface is colder and the interior of the soil is hot. We have this situation due to continentality and high temperature (about 35-40 ºC). When you compare these measures with sun exposed soil and this one protected by a small shade (shrubs or herbaceous plants) the difference is more significant. I think these kind of data are not possible of obtaining by satelite images or any indexes.

Several studies have investigated the relationship between surface and root-zone measurements. You can easily find them on main hydrological Journals.

For the validation of remote sensing products, sensors installed horizontally at 5 cm depth are used for practical reasons, i.e., it is considered the sensor closer to the surface.

Anyhow, I would suggest you reading these two recent papers (see attached link), analyzing in depth the relationship between surface and root-zone soil moisture with novel approaches (one is from our group :)):

Article The impact of vertical measurement depth on the information ...

Article A physically based approach for the estimation of root-zone ...

I'm working on soil moisture estimation through temperature measurement using fiber optic. According to my experience, I would say that the difference between soil moisture at 2 and 4 cm of depth can be important during and just after rainfall events (0.1). This difference is larger in fine soils with low permeability. In between the rainfall events, I generally do not observe large differences but it certainly depend on your meteorological conditions.

I suggest you to simply model this phenomena using the free code Hydrus1D (http://www.pc-progress.com/en/Default.aspx?Downloads). This would allow you to get a response for different type of soil, meteorological conditions, vegetation, etc...

Dear Mateusz,

You can refer to the below source:

George P. Petropoulos, 2014, Remote Sensing of Energy Fluxes and Soil Moisture Content,CRC Press, Chapter 16 (p 391-410).

In Chapter 16, Montzka reviews the recent advancements in the assimilation of remotely sensed soil surface moisture into numerical models both for synthetic and real-world studies. He also discusses recent works focusing on the soil moisture profile estimation by assimilation of one remotely sensed soil moisture product and how recent studies deal with different scales in remote sensing soil moisture observations and numerical models. Montzka also commends on the ability of different approaches in estimating model parameters for improved predictions and reviews the impact of soil moisture data assimilation for the improved estimation of further components of the hydrological cycle.

Dear Mateusz,

You can refer to the below source:

George P. Petropoulos, 2014, Remote Sensing of Energy Fluxes and Soil Moisture Content,CRC Press, Chapter 16 (p 391-410).

In Chapter 16, Montzka reviews the recent advancements in the assimilation of remotely sensed soil surface moisture into numerical models both for synthetic and real-world studies. He also discusses recent works focusing on the soil moisture profile estimation by assimilation of one remotely sensed soil moisture product and how recent studies deal with different scales in remote sensing soil moisture observations and numerical models. Montzka also commends on the ability of different approaches in estimating model parameters for improved predictions and reviews the impact of soil moisture data assimilation for the improved estimation of further components of the hydrological cycle.

soil moisture dynamics with depth are different in behavior. and it also depends on initial soil moisture or available water contents at time X. however surface soil moisture from 1-15 cm show high variation and dependencies on prevailing environment as it directly gaining or loosing moisture through rainfall or evaporation and follows normal energy balance laws. whereas with the depth this variation do not follow the exact as near surface soil moisture dynamic. 

this phenomenon is more visible in unsaturated soils or soils in dry land. where soil moisture fluctuations largely dependent on meteorological conditions. 

with increase of depth soil moisture will higher in case of dry land. 

Dear Mateus,

I think the satelite analysis are in different scales as we do in the field. In addition, the climate of the region also influences so much in depth and depending on the hour of the day and plant cover. I will search for any publication for sending, ok?

Dear All,

thank you for all your answers, pointing that other factors such as climate and hour of the day also influences soil moisture variability as well as attached, linked articles. It is really a pleasure to read such interesting responses.

However, some answers seems to be contradictory and it was not commented.

Lucca Brocca wrote:

For the validation of remote sensing products, sensors installed horizontally at 5 cm depth are used for practical reasons

On the other hand C.J. Stigter wrote:

Comparison of "near surface soil moisture" measured remotely and "in situ" is almost impossible because both sensors integrate differently (in depth and space) and in ways that therefore cannot be compared.

I (hope that?) understand both points of view. However, even if it is not possible to "fully" compare "in-situ" and remotety sensed soil moisture, it was done in numerous articles and seems reasonable.

It's always interesting to compare the different opinions in science. What I learned is to avoid to make definitive statement ("cannot be compared").

I suggest reading the two attached papers for the validation of satellite soil moisture observations. In Wagner et al. (2014) we have discussed some of the issues in the validation of satellite soil moisture products and we have provided some guidelines.

Good luck! L

http://hal.ird.fr/ird-00828830/document

Article Clarifications on the "Comparison between SMOS, VUA, ASCAT, ...

Mateusz, this is a good question. I don't know of any work that examines the very shallow variation you are interested in but we have dabbled in 0-1+[m] profile measurements. See figure 3c in (High-frequency analysis of the complex linkage between soil CO2 fluxes, photosynthesis and environmental variables TREE PHYSIOLOGY 32:49-64 · JANUARY 2012) The link is below, and the science was rather primitive in the soil moisture world but it might lead you to some of the data you need. Many of the long term eddy covariance sites will have such profiles. Check out Ameriflux, Fluxnet for sites and data access. Good luck!

Article High-frequency analysis of the complex linkage between soil ...

How representative are remote sensing observations. The penetration depth of a radar signal depends on this frequency, but also on soil properties, soil moisture content and on the presence of vegetation.  For example a C-band SAR  has a wavelength of 5.5 cm, corresponding to a penetration depth of the order of 2 cm, maximum 5 cm.

I'm currently working on a research project where we used 5 cm depth ground observation to validate the radar product. This for practical reasons. Having sensors buried at 2 cm is on easy, they can easily get damaged and if you have vegetation at 2 cm you have a lot or roots.

We have an experimental site in Alpine grasslands where for this purpose we installed sensors at 2, 5 and 20 cm depth.  Usually 2 cm and 5 cm are well correlated, but differences may happen especially after light rain events

To understand better the impact of depth on radar signal this paper could be interesting: 

Zribi, M., Gorrab, A., Baghdadi, N., Lili-Chabaane, Z., Mougenot, B., 2014. Influence of radar frequency on the relationship between bare surface soil moisture vertical profile and radar backscatter. IEEE Geosci. Remote Sens. Lett. 11 (4).

If somebody is interesting to analyze such a data please contact me. To have a look on the region please look to the attached publication.

Article Influence of Radar Frequency on the Relationship Between Bar...

Article Estimation of soil moisture patterns in mountain grasslands ...

Hi All,

As all of the contrubuters to this post has talked about MW SM estimation and its depth...i would like to add anothor option, not directly related to your question.

SM estimation using TIR such as done by the ALEXI (atmospheric land exchnage inverse model) can provide a composite SM estimate from surface to the effective rooting depth (ranges from top cm to 200 cm or more). This is another option and relatiely more suitable for agrcultural aplication.

We have done some studies regarding TIR based SM, developed a profile using the principle of maximum entropy and applied it to an agriculatural model with some success. please refer to my papers for more details..

In short, TIR based SM estimations can be compared with in-situ observations and has relatively greater applicability in agricultural fields since it provides estimates upto rooting depth not just surface.    

Dear, Actually soil moisture variation in particular at the surface soil is depending upon the conservation practices adopted viz. Mulching. Further its rate and mode effect the moisture dynamics significantly.

Further, dear friend as per our studies, the soil moisture dynamics sometimes entirely changed during the intervening period as during cropping season mulched plots had higher moisture content but during the intervening period mulched plots reported to have higher evaporation losses. you could assess my account on research gate for further information.

Regards

I still remain on scale subject.

I  am fine with the scale problem, that is an issue.

But also there is horizontal inhomogeneity.

Quite some time ago we published for irrigated fields in the Gezira, Sudan:

A.A. Ibrahim, C.J. Stigter, A.M. Adeeb, H.S. Adam and W. Van Rheenen, 1999.

On-farm sampling density and correction requirements for soil moisture determination in irrigated heavy clay soils in the Gezira, Central Sudan.

Agric. Water Managem. 41: 91 - 113.

This is not available electronically, but when you send me your physical mailing address I can air mail it to you. My e-mail address is [email protected]

Best wishes, Kees