Indications of Surface and Sub-Surface Hydrologic Properties from SMAP Soil Moisture Retrievals

Paul Dirmeyer,Holly Norton

doi:10.3390/hydrology5030036

Paul Dirmeyer, Holly Norton

Open Access

https://doi.org/10.3390/hydrology5030036

Copy DOI

Abstract

Variability and covariability of land properties (soil, vegetation and subsurface geology) and remotely sensed soil moisture over the southeast and south-central U.S. are assessed. The goal is to determine whether satellite soil moisture memory contains information regarding land properties, especially the distribution karst formations below the active soil column that have a bearing on land-atmosphere feedbacks. Local (within a few tens of km) statistics of land states and soil moisture are considered to minimize the impact of climatic variations, and the local statistics are then correlated across the domain to illuminate significant relationships. There is a clear correspondence between soil moisture memory and many land properties including karst distribution. This has implications for distributed land surface modeling, which has not considered preferential water flows through geologic formations. All correspondences are found to be strongest during spring and fall, and weak during summer, when atmospheric moisture demand appears to dominate soil moisture variability. While there are significant relationships between remotely-sensed soil moisture variability and land properties, it will be a challenge to use satellite data for terrestrial parameter estimation as there is often a great deal of correlation among soil, vegetation and karst property distributions.

Highlights

Exchanges of heat, moisture, momentum, carbon, nutrients and other chemical species between land–atmosphere are important for weather, climate, hydrology, ecology, and a range of other natural and social sciences
We focus on the southeastern and south-central (SE-SC) United States in this study, where a long warm season and a great deal of heterogeneity in vegetation, soils and karst distributions exist, providing a good testbed for detection of signals from remote sensing
One problem is that weather is a confounding factor independent of land properties, for a short data set like SMAP; weather variability can lead to significant precipitation heterogeneity on scales of 10 to 100 km, contributing to similar heterogeneity in soil moisture

Summary

Introduction

Moisture, momentum, carbon, nutrients and other chemical species between land–atmosphere are important for weather, climate, hydrology, ecology, and a range of other natural and social sciences. In the aid of scientific understanding and prediction, numerical models have been developed to simulate and predict these exchanges and their effects. In weather and climate applications, land surface models (LSMs) have evolved from simple water and energy balance closure schemes into sophisticated modular codes incorporating the simulation of physical processes bridging many scientific disciplines [1,2,3,4,5]. The performance of LSMs depends on the availability of accurate data about the properties of land and their spatial distribution [6,7,8]. Vegetation is fairly well monitored from satellite [9] and on the ground, especially where related to agriculture, forestry, and other economic resources [10,11,12]

Methods

Results

Conclusion