Characterising the regional growth of on-farm storages and their implications for water resources under a changing climate

Abstract

Water demand for agricultural activities has caused a growth in small on-farm storages (i.e., ≥ 0.1 ha to ≤ 10 ha) in many areas worldwide, thereby increasing the capture of landscape runoff which otherwise would enter the catchment drainage system. The cumulative effects of these on-farm storages (colloquially known as farm ponds or farm dams) can result in reduced mean annual flows, particularly during dry years. Climate change, which will very likely exacerbate extreme weather conditions including droughts, may accelerate the growth of farm dams in semi-arid agricultural areas with limited access to surface water. The aim of this study is to evaluate farm dam development and total farm dam water volumes at regional scales using Landsat water indices (WIs). The Murray-Darling Basin (MDB, 1.061 million km2) in south and eastern Australia, which has experienced a sustained growth in farm dams in the 1990s and early 2000s, is used as a case study. Using a pixel validation dataset, four WIs were trialled for their capabilities to detect the presence of water in farm dams. Of the WIs tested here, AWEIshadow showed the best performance. As a result, AWEIshadow was used to establish the year in which water was first observed (i.e., its year of commission) in 727,081 farm dams across the MDB and to calculate the total farm dam volume in the region. Farm dams’ volume in the MDB increased more than twofold between 1990 and 2020, from 1241.0 MCM (million cubic meters) to 2,629.5 MCM. The growth differed across MDB regions, generally accelerating during the ‘Millennium Drought’ (2001–2009). In the entire MDB, there was rapid development in the 1990s with 3.1% growth. Growth was 2.5% from 2000 to 2010, tapering off to 0.6% from 2010 to 2020.In Robertson et al. (this issue), the year-by-year growth in farm dams’ volume were used in a rainfall-runoff model thereby incorporating the time-varying effects of farm dams. Robertson et al. (this issue) show that explicitly modelling farm dams improves the model’s performance relative to a traditional rainfall-runoff model that ignores the effects of farm dams. Hence, this improves the capability of assessing the direct impacts of farm dams and their potential growth in climate change studies.