Skip to content

Confirmation of Candidature - Monika Rawat

Uncertainty in Predicting Soil Water and Plant Available Water at High-Resolution Scales
When
21 SEP 2022
9.10 AM - 10.40 AM
Where
Online

Soil water also termed soil moisture (SM) is often estimated in the form of plant available water (PAW) due to its inextricable connection with crop yield. Currently, three different approaches are used for providing soil water estimates: in-situ direct observations; direct or process able products from earth observation satellites (EOS) and employing a simulation model with a soil water balance algorithm. These techniques when used individually or coupled with one another tend to generate a large number of random or systematic errors defining uncertainty in predictions of soil moisture. The causes of these uncertainties are numerous. Because in-situ observations are scarce and are uneconomical to populate uniformly representing large spatial extent, remote sensing retrievals are appropriate alternatives to compensate upon this. Moreover, remote sensing technology not only renders systematic observations at various scales but also capable of providing input to environmental models in data scarce locations. However, random inconsistencies in remotely sensed data give rise to uncertainty in soil moisture predictions. Hence, identification of these uncertainties is important for practitioners to have an insight into uncertainty sources which may prevent them from inferring incorrect and bias conclusions. 

Use of model as a reliable instrument of predictions is gaining momentum by considering uncertainties and their sources in various model communications. However, uncertainty studies barely assume a generic standard owing to their dependence on the quality of model input and a prediction friendly multi-source observation data either measured or derived products from Earth observation satellite (EOS). This research aims to evaluate and explore options to minimise uncertainties or error variances in soil moisture modelling. This will be through investigating combinations of tools and open-source geo-data products that include remotely sensed and processed data obtained from EOS while integrated into a soil-water-balance (SWB) model either as input or observation proxies. Further, the usefulness of the easily accessible EOS surrogates over cost intensive soil moisture sensors to calibrate and validate the model will, also, be investigated. The outcome envisaged from this study is expected to build a scientifically robust technique that can seamlessly simulate plant available water content (PAWC) at high spatial resolution with high degree of confidence. 

Further, this research could provide growers and decision-makers valuable information about soil water availability in root zone for taking critical decisions for their system management. It would be beneficial in near-real-time drought and crop growth analysis across a landscape, especially where rain-fed farming requires timely decisions on input management for greater crop productivity.

For more information, please contact the Graduate Research School or phone 07 4631 1088.