Don't you think this is very limited insitu information for calibration of an algorithm.

Yes, I agree with you, but unfortunately we had very limited time and budget for getting ground data.

But, did you made measurements for a single date?

If you have more sampling dates, the measurements could be enough...

Ray et al., (2010) "Landslide susceptibility mapping using downscaled AMSR-E soil moisture: A case study from Cleveland Corral, California, US" may be of some use. Their aim was different but some of their methods may be applicable for your objectives

Yep they are using MODIS

Dear Luca,

We have measurements for three different dates each for 10-20 different points.

Thanks all for their contributions.

If applicable, we might compare the result of our algorithm using MODIS with other satellite soil moisture products. Is there any special product applicable in our case?

Thanks Eric. I will study the paper and let you about the applicability of its methods.

You can find the soil moisture products freely available nowadays as replies to this question on RG:

https://www.researchgate.net/post/What_is_the_best_satellite_soil_moisture_product_for_land_surface_modelling

A method based on a “universal triangle” concept was used to retrieve soil moisture from Normalized Difference Vegetation Index (NDVI) and land surface temperature (LST) data obtained from MODIS (M. Piles, A. Camps, M. Vallossera, et al., “Downscaling SMOS-derived soil moisture using MODIS visible/infrared data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 49, no. 9, pp. 3156–3166, 2011).

Thanks Luca it appears intresting article, but we see that soil moisture and NDVI dont correlate well with soil moisture, we have special cases of higher surface hydrocarbon pollution in Kuwait that creates an issue with LST.

See publication: method for determining and statistical analysis

of temperature vegetation dryness index tvdi.

Study case kampinoski national park

Sure. Thanks for your guidance.