Contribution of soil moisture variations to high temperatures over different climatic regimes

Highlights

• We assessed the correlation between soil moisture proxies and high temperatures.

• SPEI is better suited to approximate both subsoil and topsoil water content anomalies.

• There is a significant correlation at most cases when AI ranges from 0.2 to 1.

• ENSO impacts the associations particularly in semi-arid and humid regions.

Abstract

The coupling of soil moisture to climatic variables highly depends on soil moisture variability. Therefore, the activities affecting the soil moisture availability and dynamics such as land use/cover change and agricultural practices influence land-atmosphere interactions. Using soil moisture proxies, modeling studies have proven significant coupling of soil water variations to near-surface atmospheric temperature in some regions. However, these works have mostly applied the precipitation-based indices at a specific time scale. In the present study, the concurrent and lagged associations of the soil wetness variations and the high temperature index (HTI) across different climates in Iran were evaluated using Spearman’s rank test. Standardized precipitation index (SPI) and Standardized Precipitation Evapotranspiration Index (SPEI) calculated at one-month and six-month scales were employed as soil moisture surrogates. When seasonal aridity index (AI) range was from 0.2 to 1.0 (transitional condition), SPEI1 and SPI1 were correlated significantly (p <  0.05) with HTI at 83.0 and 42.4 % of cases, respectively. However, when AI was below 0.2 (water-limited condition) or over 2.0 (energy-limited condition), the topsoil moisture indices showed an insignificant association with HTI at the majority of cases. A significant negative association existed between SPEI6 and HTI at most dry cases in which soil surface is dry but subsoil appears not to be fully depleted. Therefore, the topsoil and subsoil moisture variations are likely to affect high temperature occurrence under transitional and dry conditions, respectively. The insignificant lagged correlations found for most areas indicate poor linear relationship between high temperature occurrence and soil water dynamics in the preceding month over the study area. Furthermore, when AI exceed 0.2, ENSO-related soil drought is likely to cause a stronger soil moisture-temperature coupling. Overall, the role of evapotranspiration and soil moisture depth should be considered for assessing soil moisture dynamics and land-atmosphere interactions. Our findings may be helpful for adopting soil moisture-based adaptive measures to negate the impacts of high temperatures on agricultural products.

Introduction

The soil, land cover (i.e. vegetation cover), and the overlaying atmosphere are the main components of the land-atmosphere system (Wulfmeyer et al., 2018). The interactions among the components of this system have received attention in different disciplines such as atmospheric science, land management and hydrology (Seneviratne et al., 2010; Wulfmeyer et al., 2018). Land use/cover changes such as agricultural expansion, wetland drainage, reforestation and deforestation highly impact the strength of land-atmosphere coupling via changing soil moisture budget (Hu et al., 2019; Steyaert and Knox, 2008). Therefore, the soil-vegetation-atmosphere interactions reflect the historical land use change and pattern of croplands, and are useful for reconstructing the related datasets (Steyaert and Knox, 2008). In addition, the knowledge on the soil moisture-precipitation feedback enhances the forecast skill of precipitation and climate predictability (Guillod et al., 2015; Guo and Dirmeyer, 2013). The soil moisture and temperature feedback loops also control near-surface heat flux variations. Evapotranspiration of soil water decreases near-surface temperature which is known as evaporative cooling effect (Mueller and Seneviratne, 2012; Seneviratne et al., 2010). In other words, by partially dissipating available energy via latent heat flux, soil moisture and evapotranspiration affect near-surface microclimate. Soil water shortage, therefore, induces enhanced near-surface atmospheric temperature as there is more energy remaining for sensible heating (Herold et al., 2016; Mueller and Seneviratne, 2012). Extreme hot temperatures profoundly influence different sectors, particularly in agriculture (Nouri et al., 2017; Prasad et al., 2011). Hot temperatures as a consequence of soil water deficit worsens meteorological drought impacts on agricultural systems (Bannayan et al., 2020; Nouri and Bannayan, 2019; Paymard et al., 2019). The frequency and intensity of hot extremes are also anticipated to increase under future climate change particularly in water-limited regions (Huang et al., 2017; Wang et al., 2017). Recent literature has revealed that occurrence of high temperatures and low soil moisture availability are well correlated for some regions over the globe such as the USA (Ford and Quiring, 2014; Ford et al., 2017), east China (Meng and Shen, 2014), southeastern Europe (Hirschi et al., 2010) and Australia (Herold et al., 2016; Holmes et al., 2017). Studying soil moisture-temperature coupling provides insight into forecasting heat waves (Ford et al., 2017; Orth and Seneviratne, 2014) which is helpful to adopt appropriate measures to reduce the risks imposed by high temperatures and droughts in croplands. In most regions, interannual precipitation anomalies are triggered by global-scale climate oscillations such as the El Niño-Southern Oscillation (ENSO) (Dai and Wigley, 2000; Sun et al., 2015). Thus, the soil water variations and consequently soil moisture-temperature connection appear to be affected by this large-scale oscillation over ENSO-affected areas (Ford and Quiring, 2014; Holmes et al., 2017).

Long times series of soil moisture observations are unavailable in most parts of world. Furthermore, the soil water data are mostly point-based and highly spatially heterogeneous (Robinson et al., 2008; Sims et al., 2002). Alternatively, soil moisture can be retrieved from remote sensing data or estimated by proxies, such as the Standardized Precipitation Index (SPI). As compared with remote-sensing acquisitions, proxies can be calculated over a longer time period, which is important for robust and reliable statistical analyses and more locations based on easily available climatic data (e.g. precipitation), and are available in real time (Mueller and Seneviratne, 2012). In addition, the soil profile water variability can be better estimated based on soil moisture surrogates (e.g. SPI) relative to remote sensing retrievals which mainly represent very surface soil water dynamics (Hirschi et al., 2014). The soil moisture surrogates employed within the related literature were mostly obtained on the basis of precipitation data (Herold et al., 2016). However, since the soil water balance is not only driven by precipitation but also by other components such as evapotranspiration, application of precipitation-based proxies may introduce substantial uncertainties to the results. This is of greater significance for water-limited regions where the evapotranspiration processes greatly affect soil moisture fluctuations. In addition, one specific time scale, representing soil moisture depth, has been mainly considered. Note that the soil moisture retained at soil layers may contribute to land-atmosphere interactions differently under different climatic conditions (Hagemann and Stacke, 2014; Hirschi et al., 2014).

Iran with predominantly semi-arid and arid climates experiences increasingly frequent precipitation- and temperature-related extremes (Alizadeh-Choobari and Najafi, 2017; Rahimi and Hejabi, 2018). Such extreme events seem to have adversely affected the agriculture sector and caused a range of serious socio-economic consequences in Iran (Bannayan et al., 2011; Nouri and Bannayan, 2019; Nouri et al., 2017). To the best of our knowledge, no study has been conducted to investigate the coupling strength between the soil moisture and near-surface temperature over Iran. In addition, since the climate variability can be partially explained by ENSO events over some regions in Iran (Nazemosadat and Cordery, 2000; Nazemosadat and Ghasemi, 2004; Nouri and Homaee, 2020; Sabziparvar et al., 2011), soil water-temperature associations are expected to be impacted by ENSO. Therefore, this study is aimed to assess i) the correlation between soil water variations and high temperatures using the precipitation-based and precipitation/reference evapotranspiration-based proxies calculated at different time scales, and ii) the effects of ENSO modes on soil moisture-temperature interactions over a broad range of seasonal climatic conditions in Iran.