- Title
- Towards sub-catchment scale soil moisture prediction: a combined remote sensing and land surface modelling approach
- Creator
- Senanayake, I. P.; Yeo, I. Y.; Willgoose, G. R.; Hancock, G. R.
- Relation
- 38th Hydrology and Water Resources Symposium (HWRS 2018). Hydrology and Water Resources Symposium (HWRS 2018): Water and Communities (Melbourne 3-6 December, 2018) p. 950-962
- Publisher
- Engineers Australia
- Resource Type
- conference paper
- Date
- 2018
- Description
- High resolution soil moisture information is vital for a number of applications such as water management, hydrological and climatic modelling. The available point-scale in-situ measurements and coarse-resolution (~40 km) satellite retrievals are unable to capture the sub-basin and sub-paddock scale spatial variability of soil moisture. Improving the spatial resolution of satellite soil moisture products through downscaling is a feasible solution to this problem. Thermal data based soil moisture downscaling methods perform better in semi-arid/arid regions, i.e. most of Australia. A regression model based on the thermal inertia relationship between the diurnal temperature difference (ΔT) and daily mean soil moisture (μSM) was developed in this study to estimate soil moisture at a high spatial resolution. The soil temperature and soil moisture estimates were extracted from the land surface model (LSM)-based Global Land Data Assimilation System (GLDAS) dataset (~25 km). The GLDAS data from 2000 to 2015 over the south-eastern Australia were employed in this process. The ΔT- muSM relationship is modulated by the vegetation density. The regression algorithms were tested over a 40 x 40 km area in the Upper Hunter Region of south-eastern Australia. This area was cleared mostly by cropping and grazing, representing the typical agricultural landscape and land management of south-eastern Australia. The 1 km resolution, MODerate-resolution Imaging Spectroradiometer (MODIS) derived ΔT values were input into the regression model to estimate soil moisture at 1 km resolution. These estimates were used to downscale 40 km resolution satellite soil moisture simulates developed by using the National Airborne Field Experiment 2005 (NAFE'05). The downscaled soil moisture products were validated with high resolution airborne soil moisture observations. The comparison between the high resolution soil moisture retrievals and the downscaled products shows promising results over the study area with a root mean square error (RMSE) of 0.066 cm³/cm³. The algorithms have a good potential in developing a time record of high resolution soil moisture information over the south-eastern Australia by applying them to the Soil Moisture Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) satellite soil moisture products.
- Subject
- downscaling; GLDAS; MODIS; soil moisture; thermal inertia
- Identifier
- http://hdl.handle.net/1959.13/1402689
- Identifier
- uon:35054
- Identifier
- ISBN:9781925627183
- Language
- eng
- Reviewed
- Hits: 1092
- Visitors: 1050
- Downloads: 0
Thumbnail | File | Description | Size | Format |
---|