1. Introduction
a. Background
The environmental and atmospheric impacts of land-use/land-cover (LULC) change have led to increased inquiry into the patterns and processes of land surface–climate interactions (cf. Marsh 1874; Haines 1922; Bonan 2008); that is, the extent to which different land surface types such as crops, forest, prairie, or urban, impact local to regional climate within the context of global change (e.g., Bonan 1997; Carleton et al. 2008b; Gerken et al. 2018). In particular, growing interest in the role of vegetation type and vegetation status (e.g., phenology, leaf area index) in warm-season convective precipitation (CVP; see Table 1 for definitions of acronyms) suggests the potential for mesoscale vertical circulations to develop around vegetation boundaries, enhancing CVP (Anthes 1984; Mahfouf et al. 1987; Copeland et al. 1996; Carleton et al. 2008b). Carleton et al. (2008b) and Hiestand and Carleton (2020) find that the fluxes of sensible and latent heat (evapotranspiration) differ between cropland and remnant forest in the U.S. Corn Belt (CB), according to both phenological stage and synoptic circulation pattern type.
Commonly used abbreviations in this paper and their definitions.
Less well studied for the CB but still likely important is the role of soil moisture (SM) in CVP, given that SM influences convection-supporting factors such as atmospheric boundary layer (ABL) development, evapotranspiration, and the ratio of sensible (H) to latent heat (LE) fluxes, or Bowen ratio (Alfieri et al. 2008; Ford et al. 2015; Duerinck et al. 2016; Yang et al. 2018a). Particularly in areas lacking extensive tree cover—such as the U.S. Great Plains—dry soils can deepen the ABL, which enhances uplift and instability, increases the relative humidity at the top of the ABL, and triggers more rapid convection than moist soils (Garrett 1982; Findell and Eltahir 2003a; Ek and Holtslag 2004; Ford et al. 2015). Further, the sensitivity of sensible and latent heat fluxes to SM has been shown to increase with decreasing SM, which supports CVP over dry soils (Taylor et al. 2012). Conversely, moist soils enhance static instability by providing moisture to the atmosphere via latent heating from evaporation, which can also support CVP (Alfieri et al. 2008; Ford et al. 2015; Duerinck et al. 2016). Although the CB is characterized by a humid climate, the region lacks extensive tree cover due to deforestation for agriculture in the nineteenth century (Bonan 1997); accordingly, SM is likely to influence climatic conditions for the CB. Consequently, a lack of clarity exists around the exact role of SM in CVP development, particularly for the CB.
The previously mentioned contrasting results for dry/moist soil preferences in terms of CVP development likely are attributable to the different temporal and spatial scales used within SM–CVP investigations (Guillod et al. 2015; Chen and Dirmeyer 2017), and the coinfluences of low versus high atmospheric humidity (Rabin et al. 1990). Whereas some studies compare SM with same-day CVP (e.g., Guillod et al. 2015; Chen and Dirmeyer 2017; Welty and Zeng 2018), others relate SM to next-day CVP (e.g., Alfieri et al. 2008; Tuttle and Salvucci 2016), and still others analyze SM–CVP average relationships on a longer-term (e.g., seasonal) basis to investigate potential lag effects of soil moisture on CVP (e.g., Findell and Eltahir 1997; Duerinck et al. 2016). Additionally, the spatial domain of previous SM–CVP studies spans from large scale (planetary, continental, country) (e.g., Findell and Eltahir 2003b; Taylor et al. 2012; Tuttle and Salvucci 2016; Chen and Dirmeyer 2017) to more regional scale (e.g., Gentine et al. 2013; Ford et al. 2015; Gerken et al. 2018; Yang et al. 2018a). Accordingly, the variations in results among SM–CVP investigations highlight the need for an understanding of SM–CVP patterns and processes across a range of temporal and spatial scales, applied to different regions of varying climatic and land surface characteristics (Ek and Mahrt 1994; Findell and Eltahir 2003b).
To better understand SM influences on convection, several studies have accounted for its association with other atmospheric variables, particularly the near-surface air humidity and presence or absence of low-level jet streams (LLJs) (e.g., Rabin et al. 1990; Findell and Eltahir 2003a; Ford et al. 2015; Gerken et al. 2018). In a two-day case study for Oklahoma, Rabin et al. (1990) suggest that preferential development of convective clouds in relation to the surface SM is strongly influenced by whether the atmosphere is dry or moist. These authors find that convection develops first over dry soils during low air humidity conditions on day one, but over wet soils during high humidity conditions on day two. Somewhat similarly, a one-dimensional modeling study by Findell and Eltahir (2003a) finds preferential CVP occurrence over dry soils during low humidity, and over moist soils with medium (average) humidity. The steepened temperature lapse rates and deepened ABLs on low humidity days seem to favor CVP over dry soils, whereas reduced lapse rates and higher latent heat fluxes on medium humidity days may suppress the “dry soil advantage,” leading to more likely CVP occurrence over moist soils.
In addition to the role of near-surface humidity, long-term SM–CVP average relationships have been determined with respect to the presence or absence of an LLJ. In their 10-yr observational study for the southern Great Plains, Ford et al. (2015) find CVP to occur preferentially over dry soils with an LLJ present and over moist soils lacking an LLJ. Also, these authors find a positive correlation between low near-surface humidity and low SM and between high near-surface humidity and high SM for CVP occurrence. Low SM–LLJ present days are also associated with low near-surface humidity, supporting the findings of Rabin et al. (1990). Further, Ford et al. (2015) highlight the importance of placing SM–CVP associations into the context of synoptic near-surface atmospheric patterns, an approach supported by other land surface–atmosphere studies involving LULC, particularly for the Corn Belt (e.g., Carleton et al. 2008a; Hiestand and Carleton 2020). It is also important to note that although LLJs occur frequently for the Great Plains, they also occur for the CB (e.g., Walters et al. 2008; Doubler et al. 2015).
An important caveat to the applicability of most of the aforementioned studies on SM–CVP–LLJ associations to the U.S. Corn Belt is that the adjacent Great Plains study region experiences less annual and growing season precipitation and greater potential evapotranspiration than the CB (Carleton et al. 2008a, Fig. 1). Also, grain crop types—in particular, wheat—predominate in the Great Plains and there is an increased reliance on irrigation there. Consequently, an observational analysis of the relationships between SM, the near-surface air humidity, and low-level jet presence or absence on CVP remains to be conducted in a climatic (time averaged, spatially varying) framework for the CB.
Average daily precipitation rate (mm day−1) between May and September for the central United States from 1981 to 2010. The image was provided by the NOAA/ESRL Physical Sciences Division (http://www.esrl.noaa.gov/psd/).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Average daily precipitation rate (mm day−1) between May and September for the central United States from 1981 to 2010. The image was provided by the NOAA/ESRL Physical Sciences Division (http://www.esrl.noaa.gov/psd/).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Average daily precipitation rate (mm day−1) between May and September for the central United States from 1981 to 2010. The image was provided by the NOAA/ESRL Physical Sciences Division (http://www.esrl.noaa.gov/psd/).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
The CB is a close-to-ideal location to study land–atmosphere interactions due to the limited topographic relief and presence of widespread, predominantly rain-fed (Fig. 1) croplands surrounding remnant deciduous forest and prairie land-cover types that amplify the effects of surface vegetation on climatic conditions (Carleton et al. 2001, 2008a). The intensive cultivation of these water-loving crops enhances the latent heat flux during the peak of the warm, or growing, season due to their low stomatal resistance (Bonan 1997; Sacks and Kucharik 2011; Hiestand and Carleton 2020). This increased evapotranspiration, in turn, raises the ABL humidity and summertime heat indices (e.g., Hill et al. 2019). Additionally, the extensive agriculture in the CB is supported by nutrient-rich alfisol and mollisol soil types, both of which have high moisture capacity (Soil Survey Staff 1999; Matyas and Carleton 2010).
b. Research hypotheses and paper structure
The present study determines the categorical associations of SM, the near-surface air humidity and 850-hPa winds (for detection of an LLJ) in the CB, to determine their mutual influences on CVP for the nine warm seasons (i.e., 1 May–30 September) of 2011–19. Each warm season is split into three subseasons (early, middle, and late) to account for vegetation phenology stages (e.g., Carleton et al. 2008b; Matyas and Carleton 2010). Statistical associations among those variables are identified and evaluated within the context of synoptic circulation conditions (e.g., 300-hPa wind for upper-tropospheric jet, sea level pressure). These associations determine the larger-scale atmospheric controls on CVP and their likely cross-scale interactions, from which the processes and potential mechanisms associated with significant SM–near-surface humidity–LLJ combinations and CVP may be inferred (e.g., Carleton et al. 2008a; Matyas and Carleton 2010). Based on the previously detected positive SM–humidity relationships (dry SM and low humidity; moist SM and high humidity) for CVP in the neighboring Great Plains region (Rabin et al. 1990; Findell and Eltahir 2003b), in addition to the previous findings of negative SM–LLJ associations for CVP for the Great Plains (Ford et al. 2015), we hypothesize that in the presence of an LLJ, Corn Belt CVP will occur more frequently over drier soils for lower-humidity conditions, and less frequently over wetter soils for higher-humidity conditions. Absent an LLJ, we hypothesize that CVP will occur more frequently over wetter soils for higher-humidity conditions and less frequently over drier soils for lower-humidity conditions. For dry surface conditions, the inferred strong sensible heat fluxes are expected to support CVP when coupled with moisture advection via an LLJ; conversely, moist surface conditions are expected to support CVP on days lacking strong moisture advection via an LLJ.
The rest of this paper is organized as follows: section 2 describes the data and methods of analysis, while section 3 presents the results. Discussion of those results and their implications for climate interactions with rain-fed agriculture in the CB and other midlatitude humid lowland regions follow in section 4. Section 5 presents the study’s summary and concluding remarks.
2. Data and methods
a. Study region and period of analysis
The CB study region consists of the states of Iowa, Illinois, Indiana, Ohio, and parts of the adjacent states of Michigan, Wisconsin, Minnesota, Nebraska, and Missouri (Fig. 2). We use data from four soil climate monitoring stations within the CB for the nine consecutive warm seasons of 2011–19. This 9-yr period spans a range of meteorological conditions, permitting a reasonably comprehensive assessment of SM–CVP associations, especially given that recent research suggests a 5–7-yr length of record is generally required to generate SM dataset stability (Ford et al. 2016; Leeper et al. 2019), with shorter time periods required for shallower (e.g., 5–10-cm soil depth) observations (Ford et al. 2016). For this reason, we use the 4-in. (approximately 10 cm)-depth SM data; additionally, using top-layer soil moisture data allows for more viable comparisons of our results with those of other recent SM–precipitation coupling studies that have used top-layer data (e.g., Frye and Mote 2010; Ford et al. 2015). Further, Ford et al. (2016) note that shorter periods of record tend to be required for spring and summer SM data due to the reduced potential for error from frozen soil water content during the warm season relative to the winter. Accordingly, our nine-warm-season study should satisfy requirements for dataset stability. Although an even longer time period might be ideal, the scarcity of long-term, spatially extensive soil moisture datasets exerts a temporal constraint on the analysis (Duerinck et al. 2016; Leeper et al. 2019).
Land covers for the Corn Belt, with the four SCAN stations used for this study indicated by large blue circles; three additional Corn Belt SCAN stations are indicated by smaller black circles (Yang et al. 2018b).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Land covers for the Corn Belt, with the four SCAN stations used for this study indicated by large blue circles; three additional Corn Belt SCAN stations are indicated by smaller black circles (Yang et al. 2018b).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Land covers for the Corn Belt, with the four SCAN stations used for this study indicated by large blue circles; three additional Corn Belt SCAN stations are indicated by smaller black circles (Yang et al. 2018b).
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
b. Data
The observed data on the 4-in.-depth SM and near-surface atmospheric vapor pressure (VP; utilized as a proxy for specific humidity; e.g., Wesely 1976) acquired daily at 0800 LST (1300 UTC for eastern time and 1400 UTC for central time), the daily total precipitation, and the 850-hPa wind velocity (speed and direction; V 850) observed at 1200 UTC, are obtained for the nine warm seasons (i.e., 45 months in total). The SM and VP data are provided by the Soil Climate Analysis Network (SCAN) and are each organized into categories of low (lowest third of SM or VP values), medium (middle third of SM or VP values), and high (highest third of SM or VP values) (see appendix A). The SCAN provides near-surface and subsurface climate data for over 200 stations in the United States, with a focus on agriculture and drought monitoring (Schaefer et al. 2007).
The four SCAN stations chosen to assess the Corn Belt intraregional variability in soil moisture–convective precipitation (SM–CVP) relationships are located in west-central Ohio (OH), northwestern Iowa (IA), northern Missouri (MO), and southeastern Nebraska (NE) (filled blue circles on Fig. 2). In contrast to the neighboring Great Plains region, where annual precipitation P is less than annual evaporation E, all four of these stations are located in the humid Corn Belt domain, where P > E. Stations are chosen based on the following attributes: 1) a long record spanning the study period, 2) a minimal amount of missing data, and 3) the broader physical characteristics (e.g., LULC, soil-type characteristics) of the site. Land covers surrounding the four stations are primarily croplands, with the MO station alone situated among a mixture of pasture, grasses, and crops (Fig. 2). Three other Corn Belt SCAN stations in central IA, central Illinois, and southwestern Wisconsin (filled black circles on Fig. 2) are not considered for this analysis because they do not meet the three criteria listed above.
Three-hourly accumulated precipitation data obtained from the North American Regional Reanalysis (NARR) are combined into 24-h totals (Mesinger et al. 2006). Precipitation totals are calculated from 1200 to 1200 UTC (0700–0700 LST) for the OH station and from 1500 to 1500 UTC (0900–0900 LST) for the IA, MO, and NE stations using the respective NARR grid boxes; this method accounts for interstation time zone differences and ensures that precipitation accumulation is calculated as closely to 0800 LST (the time of SM and VP data collection) as possible. The NARR supplies reanalyzed (i.e., modeled) CVP data, to permit a distinction between CVP and nonconvective (dominantly stratiform) precipitation. We classify a CVP day as any day receiving greater than 3 kg m−2 (equivalent to 3 mm) of precipitation, a threshold used in other SM–CVP studies (e.g., Taylor et al. 2012; Ford et al. 2015). We consider the full daily (i.e., 24-h) CVP totals to incorporate both daytime and nocturnal (e.g., from mesoscale convective systems) sources of CVP for the CB. We deem any day having precipitation greater than zero but not exceeding 3 mm to be a non-convective-precipitation (NCP) day.
Three-hourly average data on V 850 data (m s−1) for 1200 UTC (i.e., the average velocity for 900–1200 UTC) are considered relative to the location of each SCAN station and are provided by Earth Science Research Laboratory (ESRL) reanalysis (Kalnay et al. 1996). The 850-hPa standard pressure surface is appropriate for portraying the lower-tropospheric free-atmosphere airflow in the CB region, as lower levels (e.g., 925 hPa) are likely subject to surface friction inducing nonrepresentative conditions (e.g., Anav et al. 2010). Through visual analysis of the mean V 850 at 1200 UTC (0600/0700 LST), we determine the presence or absence of an LLJ. An LLJ is deemed present if V 850 1) is from a southeasterly, southerly, or southwesterly direction and 2) has a magnitude of at least 11 m s−1 over a station’s location as represented by the ESRL grid. The broad southerly component requirement distinguishes LLJs likely to be advecting moisture from the Gulf of Mexico (Bonner 1968; Walters et al. 2008). The moisture advection by these southerly LLJs is associated with increased buoyancy and greater static instability, as is evident through the convective available potential energy (CAPE) index (e.g., Frye and Mote 2010). The 11 m s−1 threshold is lower than in some recent SM–precipitation studies for the adjacent Great Plains region (e.g., Frye and Mote 2010; Ford et al. 2015) because average southerly and southwesterly LLJ wind speeds are lower for the CB region (10–15 m s−1) than the Great Plains (15–20 m s−1) (Bonner 1968; Walters et al. 2008).
To corroborate the accuracy of this LLJ identification process, average wind speeds and standard deviations by pressure level (500, 700, 850, 925, and 1000 hPa) are determined using 1200 UTC sounding data from the University of Wyoming (University of Wyoming 2021) for the 2011 warm season (Table 2). Sounding data are compared with 1) OH station LLJ results, using Wilmington, OH (KILN), sounding data (about 40 miles south of the OH station), and 2) NE station LLJ results, using Omaha, NE (KOAX), sounding data (about 30 miles north of the NE station). Sounding results for both KILN (OH) and KOAX (NE) indicate that mean 850-hPa wind speeds are well above the 11 m s−1 threshold, demonstrating the accuracy of our LLJ identification process.
The 1200 UTC wind speed mean and standard deviation (SD) for select pressure levels for Wilmington(KILN) and Omaha (KOAX), on LLJ days during the 2011 warm season.
Frequency data (number of days) on SM and VP categories are determined for the whole growing season and for the three subseasonal groups (early, middle, late) each of equal length (51 days each) for all nine years considered together. The division into subseasons is intended to capture the broad variation in vegetation phenology for the CB, defined after Carleton et al. (2008b) and Hiestand and Carleton (2020), as follows: 1 May–20 June (“early season”), 21 June–10 August (“midseason”), and 11 August–30 September (“late season”). For each SCAN station, any given subseason dataset having more than five days of missing data (i.e., 10%) is removed from consideration for that station. If more than two subseasons are removed from the 9-yr subseason dataset for any station, both the subseason and the respective whole-season datasets for that station are not considered further. Upon assessing the amount of missing data, the IA whole season and late season were removed from the analysis, although the early and midseason data were retained. Consequently, 14 station-season datasets remain for the analysis of SM, VP, and LLJ associations, as follows: OH-Whole, OH-Early, OH-Mid, OH-Late, IA-Early, IA-Mid, MO-Whole, MO-Early, MO-Mid, MO-Late, NE-Whole, NE-Early, NE-Mid, and NE-Late.
c. Determination of SM–VP–LLJ joint associations
To determine joint associations of SM, VP, and LLJ for the Corn Belt, the daily departure of each variable from its 9-yr (growing season, subseason) average is first calculated. The station daily SM and VP values are grouped into percentiles with respect to each season’s full-season and subseason datasets (e.g., the percentiles for OH-Whole are different from percentiles for OH-Early), resulting in 9-yr “climatologies” of SM and VP for the whole growing season and the three subseasons for each station. The association of CVP occurrence with SM, VP, and LLJ, is determined from the construction of contingency tables (nine columns of different SM–VP combinations for both LLJ and no-LLJ) for each seasonal and subseasonal group. Additionally, the percentage of CVP days with—versus without—an LLJ (relative to all CVP days) is calculated for each dataset and for each SM–VP–LLJ joint association.
Statistical significance testing applied to the abovementioned contingency tables identifies those stations and their seasonal and subseasonal groups (e.g., OH-Whole, OH-Early, etc.) likely to exhibit greater frequencies of SM–VP–LLJ configurations associated with CVP; specifically, Fisher’s exact testing is applied to the CVP frequency data for each of the 14 tables. Because some joint association categories have low frequencies (i.e., <5 days), this test is employed instead of more common statistical tests for categorical data, such as chi square, due to the lack of a minimum frequency requirement within categories for the Fisher’s test. The resulting p values reveal the likelihood that the frequency distribution of CVP and NCP for the various SM, VP and LLJ combinations did not occur by chance at the 0.05 level (i.e., 95% confidence level). If a result did not occur by chance, then this implies that the joint association is likely to be meaningful for the study period.
d. Composite mapping of tropospheric fields
To depict the atmospheric environments and synoptic features associated with different SM, VP and LLJ combinations, we employ composite mapping of daily averaged mapped fields from the NCEP–NCAR reanalysis project (e.g., Carleton et al. 2008a; Matyas and Carleton 2010). NCEP–NCAR reanalysis data are provided by the ESRL Physical Sciences Division (Earth Science Research Laboratory 2020; Kalnay et al. 1996). The CB study region is divided into two broad longitudinal sections for composite map analysis—the eastern Corn Belt (ECB) and western Corn Belt (WCB)—to permit an intraregional comparison of SM–CVP synoptic associations. Two stations each representative of ECB (OH station) and WCB (NE station) are chosen for the intraregional comparisons considering 1) their locations on opposite sides of the CB and 2) a minimal amount of missing data. Based on statistical significance testing and SM–VP–LLJ combinations of greatest frequency for these two stations, the composite map analysis is conducted for the ECB and WCB for the following variables: 300-hPa vector wind V 300, 500-hPa geopotential height Z 500, V 850, 850-hPa specific humidity, 1000-hPa specific humidity, 0–10-cm soil moisture fraction, sea level pressure (SLP), and CAPE.
3. Results
a. SM and VP cross-station comparisons
First, to assess the spatial representativeness of the four SCAN stations used in this study, cross-station correlations of SM and VP values were determined using Pearson’s product moment correlation (Table 3). The coefficients are derived using 10-day moving averages across the nine warm seasons. As would be expected, nearby stations tend to exhibit stronger correlations for both SM and VP than more distant stations. Moreover, stations exhibit much stronger correlations for VP (all correlations > 0.85) than for SM (all correlations < 0.52), a result that makes intuitive sense given that VP values reflect airmass conditions (i.e., in situ and advected heat and moisture) on synoptic scales (e.g., Bernhardt and Carleton 2019), whereas SM values are influenced on regional scales by more variable land surface conditions such as soil composition and porosity (e.g., Famiglietti et al. 1998; Salvucci 1998). Time series for warm-season SM and VP (2011–19) are displayed in Fig. 3 and show that dry SM and VP conditions prevailed more generally for 2012 and 2013, whereas wetter and more humid conditions occurred during 2015 and 2016. An additional analysis of SM–VP correlations (to determine any coinfluences between the two variables) produces negligible results, demonstrating the independence of both variables (data not shown).
Cross-station Pearson product moment correlation coefficient of SM and VP for the nine warm seasons 2011–19.
Cross-station comparisons of 10-day averaged (a) SM and (b) VP values for all four stations over the nine warm seasons 2011–19.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Cross-station comparisons of 10-day averaged (a) SM and (b) VP values for all four stations over the nine warm seasons 2011–19.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Cross-station comparisons of 10-day averaged (a) SM and (b) VP values for all four stations over the nine warm seasons 2011–19.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
The spatial representativeness of the four SCAN stations is also demonstrated in a single-season analysis incorporating data from the Illinois (IL) SCAN station. Although the IL station data do not meet the criteria for inclusion within the whole 9-yr analysis, its central location within the CB and more reliable dataset near the beginning of the study period allow for an analysis of interstation SM correlations for the 2011 warm season (Table 4). Similar to the 2011–19 correlations for the four stations, interstation correlations are generally not high. However, moderate-to-high positive correlations exist between the IL station (central Corn Belt) and both the IA and MO stations (western Corn Belt). A slight positive correlation exists between the IL station and the OH station (eastern Corn Belt); this result is somewhat stronger than the relatively small interstation correlations (±) for the OH station in the four-station correlation table. Accordingly, the IL-included correlations support our interregional results based on the original four stations.
Cross-station Pearson product moment correlation coefficient of SM for the four original SCAN stations and the IL station for the 2011 warm season.
b. LLJ frequency (spatial, temporal)
For the nine growing seasons, the LLJ frequency (V 850 field) is higher for the WCB stations (IA, MO, NE) than for the ECB station (OH) (Table 5). This result is intuitive, given the climatology of central U.S. LLJs: most frequent over and near the Great Plains during the warm season (Bonner 1968). Moreover, LLJ frequency for the CB tends to peak in the early season, with a second subseasonal maximum in the late season and a minimum in the midseason (Table 5), consistent with the intraseasonal cycle of SLP synoptic types favoring southerly airflow into the Midwest (e.g., Hiestand and Carleton 2020). The NE station is an exception to this intraseasonal pattern, whereby the late season experiences the highest LLJ frequency. The 21.4% frequency of LLJ days relative to all days for NE-Late is the highest of all 14 datasets. The percentage of LLJ days also occurring on CVP days is broadly similar (30%–45%) for WCB stations, but slightly higher (40%–50%) for the OH station. This result suggests a stronger association between LLJ and CVP occurrence for the ECB than for the WCB (Table 5). Further, the WCB stations experience a maximum in LLJ days occurring with CVP during the early season, while OH (ECB) experiences a maximum frequency of LLJ days with CVP for the late season of 50%.
Percentage of all days (CVP and NCP) with LLJ present and percentage of LLJ days associated with CVP for the 14 datasets. Statistically significant (p < 0.05) location–time couplets are indicated in boldface type.
Of the 14 station-time couplets, five (OH-Early, IA-Early, MO-Mid, MO-Late and NE-Mid) exhibit statistical significance (p values < 0.05) when compared with the expected frequency of CVP based on the SM–VP–LLJ joint associations for each dataset. Statistical significance tends to coincide with the periods of maximum LLJ occurrence in the early season for OH and IA but corresponding to the period of minimum LLJ occurrence in the midseason for MO and NE (Table 5). Interestingly, the station with greatest warm-season LLJ frequency (NE) exhibits statistical significance during the midseason, when LLJs are the least frequent. This result demonstrates the heightened influence of LLJ presence or absence, in the context of SM and VP status, on CVP during the midseason for the NE station.
c. SM–VP–LLJ joint configurations
1) CVP–LLJ associations
Figures 4a–d shows the frequency distribution of CVP–LLJ days with respect to SM–VP combinations. For the whole season and subseasonal groups averaged across all stations, high- and medium-VP associations make up most of the CVP–LLJ days (Figs. 4a–d). As shown in Table 6, the CVP–LLJ frequency decreases along a west–east gradient for the whole season and subseasonal groups, with the highest frequency for NE-Whole and the lowest frequency for OH-Whole. High- and medium-VP combinations occur most frequently when considered for the whole season, with high SM–high VP days occurring more frequently than any other association (Fig. 4a).
Frequency of each CVP–LLJ configuration for CVP days with respect to SM–VP category (e.g., high SM–high VP), considered for all four stations combined during the (a) whole season, (b) early season, (c) midseason, and (d) late season.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Frequency of each CVP–LLJ configuration for CVP days with respect to SM–VP category (e.g., high SM–high VP), considered for all four stations combined during the (a) whole season, (b) early season, (c) midseason, and (d) late season.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Frequency of each CVP–LLJ configuration for CVP days with respect to SM–VP category (e.g., high SM–high VP), considered for all four stations combined during the (a) whole season, (b) early season, (c) midseason, and (d) late season.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Percent frequency of each SM–VP category for CVP–LLJ days with respect to all CVP days for all datasets. Statistically significant (p < 0.05) datasets are indicated in boldface type.
Of the three subseason periods, CVP–LLJ co-occurrences occur most frequently for the early season at all stations except NE (Table 6). The low SM–high VP association occurs approximately twice as often as high SM–high VP in the early season for all stations except for that in MO. Low SM–high VP tends to be the most frequent combination for the early season, with low SM associations collectively accounting for the most CVP at each station (Fig. 4b, Table 6).
During the midseason, CVP–LLJ days occur infrequently in comparison with the two other subseasons for all stations analyzed, particularly OH (Fig. 4c, Table 6). Additionally, the CVP–LLJ frequency increases for the CB during the late season relative to the midseason. High SM–high VP is the most frequently occurring combination for the late season, although the WCB also experiences relatively high frequencies of low SM–high VP and low SM-medium VP associations during this time (Fig. 4d, Table 6).
2) CVP–NO LLJ associations
Figures 5a–d show the frequency distribution of CVP–no LLJ days with respect to SM–VP combinations. As with the category CVP–LLJ present, the high- and medium-VP categories occur most frequently across whole season and all subseasons on CVP–no LLJ days (Fig. 5). However, in contrast to CVP–LLJ, the CVP–no LLJ frequencies for the whole season are distributed across an increasing west–east gradient, with OH-Whole experiencing the highest and NE-Whole experiencing the lowest frequencies (Table 7). Additionally, across the subseasonal groups, the midseason tends to experience the greatest frequency of CVP–no LLJ days for all stations. Frequency distributions of CVP–no LLJ days are generally proportionate across the three whole-season datasets, except OH-Whole, which exhibits relatively frequent high SM–high VP occurrence.
As in Fig. 4, but for each CVP–no LLJ configuration.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
As in Fig. 4, but for each CVP–no LLJ configuration.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
As in Fig. 4, but for each CVP–no LLJ configuration.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
The CVP–no LLJ category tends to occur least often during the early season, except for NE where early- and late-season frequencies are comparable (Fig. 5b, Table 7). Considering SM–VP combinations, the low SM–high VP association generally occurs most frequently during the early season for the entire region but particularly for IA, while OH-Early experiences over 30% of all CVP days as high SM–high VP or high SM–medium VP. In contrast, the low VP associations make up less than 10% of all CVP days for the early season.
d. Spatial composites of atmospheric fields
1) CVP–LLJ combination
Spatial composites of atmospheric fields and features for CVP–LLJ days (Fig. 4) are generated separately for the early and late subseasons, which see the most frequent occurrences of CVP–LLJ. Low SM–high VP days are included for the early season, as this category occurs most often when averaged across all four stations (Fig. 4b). Conversely, because high SM–high VP days are most frequent in the late season (Fig. 4d), these composites are also presented as schematic maps depicting the average state of the surface and atmospheric variables for the CB. The composite maps depict the six variables described in the data and methods section for the OH station (to represent the ECB) and the NE station (to represent the WCB), and the presentation borrows from that given in Carleton et al. (2008a) and Matyas and Carleton (2010).
(i) Early season
For CVP–LLJ days having low SM–high VP conditions (Fig. 4; Table 6), the 300-hPa wind speeds are highest (i.e., an upper-tropospheric jet maximum) north of the CB, with both stations located due south of the right entrance portion of jet streaks, associated with upper-tropospheric divergence and thus, near-surface convergence (Figs. 6a,b). In contrast, 850-hPa wind speeds are strongest above both the ECB and WCB on these days. Considering the 500-hPa level, a trough is situated upstream of the ECB and a ridge is located downstream of the WCB–patterns associated with rising air from the surface instability and a high potential for convective precipitation (e.g., Carleton et al. 2008a).
Composite schematic representations of synoptic circulation features for (top) CVP–LLJ low SM–high VP days in the early season and (bottom) CVP–LLJ high SM–high VP days in the late season for (a),(c) ECB and (b),(d) WCB. Locations of the OH and NE station are denoted by filled black circles.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Composite schematic representations of synoptic circulation features for (top) CVP–LLJ low SM–high VP days in the early season and (bottom) CVP–LLJ high SM–high VP days in the late season for (a),(c) ECB and (b),(d) WCB. Locations of the OH and NE station are denoted by filled black circles.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
Composite schematic representations of synoptic circulation features for (top) CVP–LLJ low SM–high VP days in the early season and (bottom) CVP–LLJ high SM–high VP days in the late season for (a),(c) ECB and (b),(d) WCB. Locations of the OH and NE station are denoted by filled black circles.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
In context of the low SM–high VP association, the highest 850- and 1000-hPa specific humidity (SH) areas are located vertically above the CB on CVP–LLJ days (Figs. 6a,b). These high-SH areas coincide with the location of the LLJ, indicating the advection of humid air from the direction of the Gulf of Mexico. High CAPE values are present for the WCB, although the highest CAPE is displaced to the southwest for the ECB, located in proximity to a 500-hPa trough axis. High CAPE and moist lower-tropospheric conditions for the WCB are highly supportive of CVP (Yin et al. 2015).
On CVP–LLJ days accompanying low SM–high VP conditions for both the ECB and WCB, a broad anticyclone near the U.S. mid-Atlantic coast, and low pressure over the Great Plains, are conducive to southerly to southwesterly flow over the CB, which typically advects heat and moisture from the Gulf of Mexico. For the ECB, the OH station is situated between an area of higher SM over the western portion of the CB—and lower SM to the east (Fig. 6a). For the WCB, the highest soil moisture values are situated just to the northeast of the NE station, with the lowest values to the southwest (Fig. 6b). Consequently, the NE station is located along a relatively steep gradient between the areas of highest and lowest SM—conditions that have been shown, via modeling, to support mesoscale circulations in the boundary layer and enhanced precipitation (Cioni and Hohenegger 2018).
(ii) Late season
On CVP–LLJ days with high SM–high VP conditions (Fig. 4; Table 6), the strongest 300-hPa winds are located to the northwest of the CB (Figs. 6c,d). Both stations (OH for ECB, NE for WCB) are located close to the right entrance portion of the jet streak with attendant dynamically rising air (instability), similar to the early-season composites. Also, as in the early season, LLJs are located vertically above both stations on these days; moreover, both the ECB and WCB are situated between an upstream trough and a downstream ridge at 500-hPa, indicating divergence aloft and ascending air.
In context of these high SM–high VP associations, the low-altitude atmospheric moisture maxima are located vertically above each of the two representative stations, approximating the axis of the LLJ, particularly for the WCB. For the ECB, an additional area of higher 850-hPa SH is located over the southern Great Plains, upwind of the entrance region of the LLJ. This pattern indicates the potential for advection of moisture from the region of high 850-hPa SH into the ECB. Additionally, a broad area of high SH at 1000 hPa is present over the CB on these days. Considering the WCB, high CAPE values lie directly over the NE station and largely overlap with high-SH regions at 850 and 1000 hPa, as well as the LLJ, indicating unstable conditions.
The associated pattern of SLP is similar to that for the early season, again favoring the advection of heat and moisture from the Gulf of Mexico into the CB. Also similar to the early-season composites for the WCB (Fig. 6b), the spatial orientation of SM—lower values westward, higher values eastward—places the NE station within a narrow corridor between regions of highest and lowest SM, which has been shown to enhance CVP potential (Cioni and Hohenegger 2018).
2) CVP–NO LLJ combination
For comparison purposes, the CVP–no LLJ day map composites of the midseason are presented as the most commonly occurring subseason for this configuration. In particular, the medium SM–high VP combination is composited, which occurs most frequently when considering all four stations (Fig. 5c).
On CVP–no LLJ days having medium SM–high VP, the 300-hPa vector wind maximum occurs to the north of the CB (Figs. 7a,b). A 300-hPa jet streak is located to the north of the OH station for the ECB, whereas the NE station is located near the right entrance portion of a 300-hPa jet streak for the WCB (i.e., increased upward motion of air). An area of relatively high (6–7 m s−1) southwesterly winds at 850-hPa that do not meet LLJ speed criteria is nonetheless present over the OH station for the ECB, whereas 850-hPa maximum winds of <11 m s−1 also are confined to the southern Great Plains for the WCB composite.
As in Fig. 6, but for CVP–no LLJ medium SM–high VP days in the midseason for (a) ECB and (b) WCB.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
As in Fig. 6, but for CVP–no LLJ medium SM–high VP days in the midseason for (a) ECB and (b) WCB.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
As in Fig. 6, but for CVP–no LLJ medium SM–high VP days in the midseason for (a) ECB and (b) WCB.
Citation: Journal of Applied Meteorology and Climatology 60, 12; 10.1175/JAMC-D-20-0285.1
A moist lower troposphere (i.e., high 850- and 1000-hPa SH) is either present or close to the CB (Fig. 7). However, in contrast to the CVP–LLJ composites (Fig. 6), high CAPE is located at both the ECB and the WCB stations. This pattern indicates unstable lower-tropospheric conditions conducive to upward vertical motion and convective precipitation for both the ECB and WCB, despite the lack of an LLJ.
The average SLP pattern for medium SM–high VP conditions on CVP–no LLJ days consists of a center of high pressure to the southeast, implying southwesterly near-surface flow. This synoptic pattern favors warm advection for both the ECB and the WCB, in addition to likely moisture advection into the ECB evident in the broad region of high 1000-hPa SH overlying high SM that is located southwest of the ECB (Fig. 7).
4. Discussion
The near-surface humidity (i.e., VP) appears to exert the most control on the occurrence of Corn Belt convective precipitation (CVP): CVP occurs more frequently on high VP days of varying SM and LLJ conditions, but on a negligible number of days having low VP. This finding agrees with Findell and Eltahir (2003a). Although past research for the neighboring Great Plains indicates a closer statistical association between CVP and the surface fluxes of sensible and latent heat (e.g., Rabin et al. 1990; Ford et al. 2015; Chen and Dirmeyer 2017), the SM–CVP coupling can also be amplified in those drier regions relative to humid regions such as the CB due to increased sensitivity of the surface heat and moisture fluxes to decreasing soil moisture (Taylor et al. 2012; Guillod et al. 2015; Yang et al. 2018a). Considering the minimal variation in CVP occurrence with respect to SM condition shown in our study, the SM–CVP coupling appears to be less pronounced for the CB than for the Great Plains.
An unexpected yet physically explainable result is that although LLJs occur much less often in the ECB than the WCB, the percentage of LLJ days accompanying CVP is higher for the ECB (45%–50%) than for the WCB (30%–45%) (Table 5). This result implies that the occurrence of CVP may be more sensitive to LLJ presence or absence for the ECB than the WCB. Although previous studies for the WCB have determined greater LLJ-associated precipitation there relative to the ECB (e.g., Jiang et al. 2007; Wang and Chen 2009), those investigations do not consider the percentage of LLJ days relative to all days experiencing CVP. Further, the sensitivity of CVP to LLJ presence is important in the context of the springtime (defined as April, May, and June) mesoscale convective system (MCS)-derived rainfall for the Midwest increasing over recent decades, which is strongly influenced by LLJs (Feng et al. 2016). The present results suggest the importance of spatial discontinuities influencing LLJ occurrence and location—and potentially also the SM—on CVP to more accurately predict future changes to MCS-derived as well as other organized (e.g., linear, squall line) rainfall for the Midwest.
The low number of CVP days co-occurring with low SM-low VP conditions for the warm season (cf. Rabin et al. 1990), may be due partly to the climatic approach of this paper, in contrast to previous case study (Great Plains) approaches: the associations suggested by Rabin et al. (1990) span a two-day period of contrasting atmospheric moisture conditions and do not account for variations in warm-season conditions considered over multiple years. Additionally, differences in climate and related LULC likely impact the contrasting results between the CB and the Great Plains; the CB contains widespread corn and soybean crop types that are rain fed, as well as remnant forest land cover, whereas wheat is more commonly grown in the drier Great Plains (e.g., Carleton and O’Neal 1995).
In finding a lack of a statistical difference in the frequency distribution of SM–VP associations between CVP–LLJ and CVP–no LLJ days, the present study also contrasts with the findings of Ford et al. (2015) for Oklahoma. In the CB, CVP occurs on both LLJ and no-LLJ days having average to above-average near-surface humidity. This lack of distinction in CVP according to LLJ presence or absence is probably explained by the lower available soil and near-surface moisture for the Great Plains contrasted with the CB, meaning that moisture advection by the LLJ is likely more crucial for CVP in the former region (e.g., Frye and Mote 2010).
Notwithstanding the demonstrated primary influence of VP on CVP in the Corn Belt, our results also suggest that SM conditions and LLJ presence or absence play additional roles in CVP according to growing-season subperiod (i.e., early, mid-, late, or whole) and location (i.e., eastern vs western CB). Low SM–high VP conditions appear conducive to CVP occurrence in the early season both with and without an LLJ necessarily being present (Figs. 4b, 5b); the CVP frequency of low SM–high VP conditions on early-season CVP–LLJ days is particularly noticeable in comparison with other CVP–LLJ days having different SM–VP conditions. An explanation for this high frequency is that crops—especially corn and soybeans—are usually planted at the beginning of the early season for the CB, meaning much bare soil is exposed prior to crop development (Carleton et al. 1994; USDA 1997; Adegoke and Carleton 2002; Carleton et al. 2008b). Thus, the lack of vegetation cover combined with particularly dry soils produces strong sensible heat fluxes (e.g., Hiestand and Carleton 2020), with accompanying static instability and ascent of air that encourage CVP (Giorgi et al. 1996; Ford et al. 2015). Additionally, dry soils can be conducive to moist convective processes given sufficient humidity near the top of the ABL; for example, that transported via an LLJ (e.g., Ek and Holtslag 2004).
A possible explanation for the generally high SM preference of CVP during the early season at the MO station—characterized by a grassland—is the low evapotranspiration rates for grasslands (e.g., Bonan 1997) in comparison with evapotranspiration for croplands (i.e., at the OH, IA, and NE stations) as crops emerge near the end of the early season. This interstation difference in evapotranspiration means that, as crops grow and their leaf area index increases, croplands will have greater associated latent heat fluxes than grasslands on low SM days (e.g., Adegoke and Carleton 2002), helping support deep convection (Bonan 1997; USDA 1997; Hiestand and Carleton 2020). Additionally, the tight longitudinal gradient in SM over the WCB on CVP–LLJ days having low SM–high VP during the early season likely enhances the potential for mesoscale circulations within the ABL, which increases CVP along these SM gradients (Fig. 6b; Cioni and Hohenegger 2018). This increased potential for deep convection may be evidenced by the presence of a region of high CAPE over much of the WCB.
Despite the decreased frontal cyclonic activity for the midseason relative to the early and late seasons, synoptic conditions still exert a considerable influence on these medium SM–high VP days; importantly, moisture advection is suggested from areas of high specific humidity at 850 hPa into the CB via non-LLJ winds (6–7 m s−1). The synoptic circulations on these midseason days, in addition to the increasing latent heat flux over maturing croplands, suggest favorable conditions for CVP.
The importance of at least adequate moisture at the surface and in the atmosphere to support deep convective activity is evident from the frequent occurrence of CVP days with high SM–high VP conditions during the late growing season. This reliance on high moisture supply for CVP—indicated by the spatial composites (Figs. 6c,d)—is intuitive, given that Corn Belt SM and the total (i.e., convective plus stratiform) precipitation totals tend to be lowest toward the end of the warm season (figures not shown).
5. Summary and conclusions
For the Corn Belt (CB) region of the Midwest United States, this study determined the statistical and spatial associations—and inferred likely physical mechanisms—of soil moisture (SM), near-surface humidity (VP), and low-level jet (LLJ) presence/absence, on convective precipitation (CVP) for the nine growing seasons of 2011–19. In situ daily SM and VP data at four SCAN stations having minimal missing data and representative of the broader physical characteristics of the CB, together with reanalysis data on LLJ and CVP presence, were used to document their climatic associations. These associations were determined via statistical testing of contingency tables (see appendix B), and construction of map composites for days exhibiting frequently co-occurring categories of SM and VP, and LLJ presence/absence.
We confirm that the joint impact of SM and lower-atmosphere humidity on CVP is different for the CB than for the more comprehensively studied Great Plains; the near-surface air humidity is shown to be the primary control on CVP for the CB, rather than SM or LLJ presence/absence. Moreover, our results suggest weaker SM–CVP coupling for the CB than for the Great Plains. As a result of these between-region differences, the present study both confirms and extends previous studies of SM–CVP coupling in the middle latitudes by including the humid lowlands of the U.S. Corn Belt.
Future work should seek to evaluate the physical processes inferred here by using a dynamical modeling approach (e.g., the WRF Model; Zaitchik et al. 2013). Subsequent research should also consider additional precipitation thresholds to determine how SM–VP–LLJ associations might vary with precipitation event intensity. Additionally, future work should consider the potential impacts of soil characteristics on SM–precipitation coupling; for example, broadscale differences in soil color, porosity and permeability related to contrasting soil types might influence boundary layer characteristics and the propensity for deep convection, especially in the soil-type transition zone (e.g., Matyas and Carleton 2010). Last, the present study should aid in better understanding land–atmosphere processes associated with warm-season precipitation and its prediction on weather time scales for the agriculturally highly productive yet climatically variable CB region.
Acknowledgments
The authors are grateful to Drs. Guido Cervone and Erica Smithwick for helpful suggestions during this research.
Data availability statement
The soil moisture and vapor pressure data are provided by the Soil Climate Analysis Network (SCAN; https://www.wcc.nrcs.usda.gov/scan/). Precipitation data are obtained via the North American Regional Reanalysis (NARR; https://psl.noaa.gov/data/gridded/data.narr.html). Three-hourly V 850 data, as well as data for spatially composited variables, are provided by the Earth Science Research Laboratory (ESRL) Physical Sciences Division (https://psl.noaa.gov/).
APPENDIX A
SM and VP Values
Table A1 shows the range of SM and VP values stratified into low, medium, and high by station and subseason.
Range of SM and VP values stratified into low, medium, and high for each station and subseason.
APPENDIX B
CVP/NCP Distributions by Station
Tables B1–B14 give the distribution of CVP and NCP days with respect to SM and VP status (low, medium, or high) and LLJ presence or absence for each station and subseason.
Distribution of CVP and NCP days with respect to SM and VP status (low/medium/high) and LLJ presence/absence (indicated as “Y” or “N”) for OH-Whole.
REFERENCES
Adegoke, J. O., and A. M. Carleton, 2002: Relations between soil moisture and satellite vegetation indices in the U.S. Corn Belt. J. Hydrometeor., 3, 395–405, https://doi.org/10.1175/1525-7541(2002)003<0395:RBSMAS>2.0.CO;2.
Alfieri, L., P. Claps, P. D’Odorico, F. Laio, and T. Over, 2008: An analysis of the soil moisture feedback on convective and stratiform precipitation. J. Hydrometeor., 9, 280–291, https://doi.org/10.1175/2007JHM863.1.
Anav, A., P. M. Ruti, V. Artale, and R. Valentini, 2010: Modelling the effects of land-cover changes on surface climate in the Mediterranean region. Climate Res., 41, 91–104, https://doi.org/10.3354/cr00841.
Anthes, R. A., 1984: Enhancement of convective precipitation by mesoscale variations in vegetative covering in semiarid regions. J. Climate Appl. Meteor., 23, 541–554, https://doi.org/10.1175/1520-0450(1984)023<0541:EOCPBM>2.0.CO;2.
Bernhardt, J., and A. M. Carleton, 2019: Comparing daily temperature averaging methods: The role of surface and atmosphere variables in determining spatial and seasonal variability. Theor. Appl. Climatol., 136, 499–512, https://doi.org/10.1007/s00704-018-2504-7.
Bonan, G. B., 1997: Effects of land use on the climate of the United States. Climatic Change, 37, 449–486, https://doi.org/10.1023/A:1005305708775.
Bonan, G. B., 2008: Ecological Climatology. Cambridge University Press, 550 pp.
Bonner, W. D., 1968: Climatology of the low-level jet. Mon. Wea. Rev., 96, 833–850, https://doi.org/10.1175/1520-0493(1968)096<0833:COTLLJ>2.0.CO;2.
Carleton, A. M., and M. O’Neal, 1995: Satellite-derived land surface climate ‘signal’ for the Midwest U.S.A. Int. J. Remote Sens., 16, 3195–3202, https://doi.org/10.1080/01431169508954623.
Carleton, A. M., D. Travis, D. Arnold, R. Brinegar, D. E. Jelinksi, and D. R. Easterling, 1994: Climatic-scale vegetation–cloud interactions during drought using satellite data. Int. J. Climatol., 14, 593–623, https://doi.org/10.1002/joc.3370140602.
Carleton, A. M., J. Adegoke, J. Allard, D. L. Arnold, and D. J. Travis, 2001: Summer season land cover—Convective cloud associations for the Midwest U.S. “Corn Belt.” Geophys. Res. Lett., 28, 1679–1682, https://doi.org/10.1029/2000GL012635.
Carleton, A. M., D. L. Arnold, D. J. Travis, S. Curran, and J. O. Adegoke, 2008a: Synoptic circulation and land surface influences on convection in the Midwest U.S. “Corn Belt” during the summers of 1999 and 2000. Part I: Composite synoptic environments. J. Climate, 21, 3389–3415, https://doi.org/10.1175/2007JCLI1578.1.
Carleton, A. M., D. J. Travis, J. O. Adegoke, D. L. Arnold, and S. Curran, 2008b: Synoptic circulation and land surface influences on convection in the Midwest U.S. “Corn Belt” during the summers of 1999 and 2000. Part II: Role of vegetation boundaries. J. Climate, 21, 3617–3641, https://doi.org/10.1175/2007JCLI1584.1.
Chen, L., and P. A. Dirmeyer, 2017: Impacts of land-use/land-cover change on afternoon precipitation over North America. J. Climate, 30, 2121–2140, https://doi.org/10.1175/JCLI-D-16-0589.1.
Cioni, G., and C. Hohenegger, 2018: A simplified model of precipitation enhancement over a heterogeneous surface. Hydrol. Earth Syst. Sci., 22, 3197–3212, https://doi.org/10.5194/hess-22-3197-2018.
Copeland, J. H., R. A. Pielke, and T. G. F. Kittel, 1996: Potential climatic impacts of vegetation change: A regional modeling study. J. Geophys. Res., 101, 7409–7418, https://doi.org/10.1029/95JD02676.
Doubler, D. L., J. A. Winkler, X. Bian, C. K. Walters, and S. Zhong, 2015: An NARR-derived climatology of southerly and northerly low-level jets over North America and coastal environs. J. Appl. Meteor. Climatol., 54, 1596–1619, https://doi.org/10.1175/JAMC-D-14-0311.1.
Duerinck, H., R. van der Ent, N. van de Giesen, G. Schoups, V. Babovic, and P. J.-F. Yeh, 2016: Observed soil moisture–precipitation feedback in Illinois: A systematic analysis over different scales. J. Hydrometeor., 17, 1645–1660, https://doi.org/10.1175/JHM-D-15-0032.1.
Earth Science Research Laboratory, 2020: NCEP North American Regional Reanalysis: NARR. NOAA, accessed 26 February 2020, https://www.esrl.noaa.gov/psd/data/gridded/data.narr.html.
Ek, M. B., and L. Mahrt, 1994: Daytime evolution of relative humidity at the boundary layer top. Mon. Wea. Rev., 122, 2710–2721, https://doi.org/10.1175/1520-0493(1994)122<2709:DEORHA>2.0.CO;2.
Ek, M. B., and A. A. M. Holtslag, 2004: Influence of soil moisture on boundary layer cloud development. J. Hydrometeor., 5, 86–99, https://doi.org/10.1175/1525-7541(2004)005<0086:IOSMOB>2.0.CO;2.
Famiglietti, J. S., J. W. Rudnicki, and M. Rodell, 1998: Variability in surface moisture content along a hillslope transect: Rattlesnake Hill, Texas. J. Hydrol., 210, 259–281, https://doi.org/10.1016/S0022-1694(98)00187-5.
Feng, Z., L.-Y. Leung, S. Hagos, R. A. Houze Jr., C. Burleyson, and K. Balaguru, 2016: More frequent intense and long-lived storms dominate the springtime trend in central U.S. rainfall. Nat. Commun., 7, 13429, https://doi.org/10.1038/ncomms13429.
Findell, K. L., and E. A. B. Eltahir, 1997: An analysis of the relationship between spring soil moisture and summer rainfall, based on direct observations from Illinois. Water Resour. Res., 33, 725–735, https://doi.org/10.1029/96WR03756.
Findell, K. L., and E. A. B. Eltahir, 2003a: Atmospheric controls on soil moisture–boundary layer interactions. Part I: Framework development. J. Hydrometeor., 4, 552–569, https://doi.org/10.1175/1525-7541(2003)004<0552:ACOSML>2.0.CO;2.
Findell, K. L., and E. A. B. Eltahir, 2003b: Atmospheric controls on soil moisture–boundary layer interactions. Part II: Feedbacks within the continental United States. J. Hydrometeor., 4, 570–583, https://doi.org/10.1175/1525-7541(2003)004<0570:ACOSML>2.0.CO;2.
Ford, T. W., A. D. Rapp, and S. M. Quiring, 2015: Does afternoon precipitation occur preferentially over dry or wet soils in Oklahoma? J. Hydrometeor., 16, 874–888, https://doi.org/10.1175/JHM-D-14-0005.1.
Ford, T. W., Q. Wang, and S. M. Quiring, 2016: The observation record length necessary to generate robust soil moisture percentiles. J. Appl. Meteor. Climatol., 55, 2131–2149, https://doi.org/10.1175/JAMC-D-16-0143.1.
Frye, J. D., and T. L. Mote, 2010: Convection initiation along soil moisture boundaries in the southern Great Plains. Mon. Wea. Rev., 138, 1140–1151, https://doi.org/10.1175/2009MWR2865.1.
Garrett, A. J., 1982: A parameter study of interactions between convective clouds, the convective boundary layer, and forested surface. Mon. Wea. Rev., 110, 1041–1059, https://doi.org/10.1175/1520-0493(1982)110<1041:APSOIB>2.0.CO;2.
Gentine, P., A. A. Holtslag, F. D’Andrea, and M. Ek, 2013: Surface and atmospheric controls on the onset of moist convection over land. J. Hydrometeor., 14, 1443–1462, https://doi.org/10.1175/JHM-D-12-0137.1.
Gerken, T., G. T. Bromley, and P. C. Stoy, 2018: Surface moistening trends in the northern North American Great Plains increase the likelihood of convective initiation. J. Hydrometeor., 19, 227–244, https://doi.org/10.1175/JHM-D-17-0117.1.
Giorgi, F., L. O. Mearns, C. Shields, and L. Mayer, 1996: A regional model study of the importance of local versus remote controls of the 1988 drought and the 1993 flood over the central United States. J. Climate, 9, 1150–1162, https://doi.org/10.1175/1520-0442(1996)009<1150:ARMSOT>2.0.CO;2.
Guillod, B. P., B. Orlowsky, D. G. Miralles, A. J. Teuling, and S. I. Seneviratne, 2015: Reconciling spatial and temporal soil moisture effects on afternoon rainfall. Nat. Commun., 6, 6443, https://doi.org/10.1038/ncomms7443.
Haines, E. H., 1922: Influence of varying soil conditions on night-air temperatures. Mon. Wea. Rev., 50, 363–366, https://doi.org/10.1175/1520-0493(1922)50<363:IOVSCO>2.0.CO;2.
Hiestand, M. P., and A. M. Carleton, 2020: Growing-season synoptic and phenological controls on heat fluxes over forest and cropland sites in the Midwest U.S. Corn Belt. J. Appl. Meteor. Climatol., 59, 381–400, https://doi.org/10.1175/JAMC-D-19-0019.1.
Hill, A. C., M. Mitchell, F. Yuan, and C. T. Ruhland, 2019: Intensification of midwestern agriculture as a regional climate modifier and atmospheric boundary layer moisture source. Ann. Amer. Assoc. Geogr., 109, 1775–1794, https://doi.org/10.1080/24694452.2019.1598842.
Jiang, X., N.-C. Lau, I. M. Held, and J. J. Ploshay, 2007: Mechanisms of the Great Plains low-level let as simulated in an AGCM. J. Atmos. Sci., 64, 532–547, https://doi.org/10.1175/JAS3847.1.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR Reanalysis 40-Year Project. Bull. Amer. Meteor. Soc., 77, 437–471, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Leeper, R. D., J. E. Bell, and M. A. Palecki, 2019: A description and evaluation of U.S. Climate Reference Network standardized soil moisture dataset. J. Appl. Meteor. Climatol., 58, 1417–1428, https://doi.org/10.1175/JAMC-D-18-0269.1.
Mahfouf, J. F., E. Richard, and P. Mascart, 1987: The influence of soil and vegetation on the development of mesoscale circulations. J. Climate Appl. Meteor., 26, 1483–1495, https://doi.org/10.1175/1520-0450(1987)026<1483:TIOSAV>2.0.CO;2.
Marsh, G. P., 1874: The Earth as Modified by Human Action. Scribner, Armstrong & Co., 656 pp.
Matyas, C. J., and A. M. Carleton, 2010: Surface radar-derived convective rainfall associations with Midwest US land surface conditions in summer seasons 1999 and 2000. Theor. Appl. Climatol., 99, 315–330, https://doi.org/10.1007/s00704-009-0144-7.
Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343–360, https://doi.org/10.1175/BAMS-87-3-343.
Rabin, R., S. Stadler, P. J. Wetzel, D. J. Stensrud, and M. Gregory, 1990: Observed effects of landscape variability on convective clouds. Bull. Amer. Meteor. Soc., 71, 272–280, https://doi.org/10.1175/1520-0477(1990)071<0272:OEOLVO>2.0.CO;2.
Sacks, W. J., and C. J. Kucharik, 2011: Crop management and phenology trends in the U.S. Corn Belt: Impacts on yields, evapotranspiration and energy balance. Agric. For. Meteor., 151, 882–894, https://doi.org/10.1016/j.agrformet.2011.02.010.
Salvucci, G., 1998: Limiting relations between soil moisture and soil texture with implications for measured, modeled and remotely sensed estimates. Geophys. Res. Lett., 25, 1757–1760, https://doi.org/10.1029/98GL01138.
Schaefer, G. L., M. H. Cosh, and T. J. Jackson, 2007: The USDA natural resources conservation service Soil Climate Analysis Network (SCAN). J. Atmos. Oceanic Technol., 24, 2073–2077, https://doi.org/10.1175/2007JTECHA930.1.
Soil Survey Staff, 1999: Soils of the United States. Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys. 2nd ed. Natural Resources Conservation Service, 837–850.
Taylor, C. M., R. A. M. de Jeu, F. Guichard, P. P. Harris, and W. A. Dorigo, 2012: Afternoon rain more likely over drier soils. Nature, 489, 423–426, https://doi.org/10.1038/nature11377.
Tuttle, S., and G. Salvucci, 2016: Empirical evidence of contrasting soil moisture–precipitation feedbacks across the United States. Science, 352, 825–828, https://doi.org/10.1126/science.aaa7185.
University of Wyoming, 2021: Sounding data. Accessed 29 April 2021, http://www.weather.uwyo.edu/upperair/sounding.html.
USDA, 1997: Usual planting and harvesting dates for U.S. field crops, accessed December 1997. National Agricultural Statistics Service Agricultural Handbook 628, 51 pp., https://swat.tamu.edu/media/90113/crops-typicalplanting-harvestingdates-by-states.pdf.
Walters, C. K., J. A. Winkler, R. P. Shadbolt, J. van Ravensway, and G. D. Bierly, 2008: A long-term climatology of southerly and northerly low-level jets for the central United States. Ann. Assoc. Amer. Geogr., 98, 521–552, https://doi.org/10.1080/00045600802046387.
Wang, S. Y., and T. C. Chen, 2009: The late-spring maximum rainfall over the U.S. central plains and the role of the low-level jet. J. Climate, 22, 4696–4709, https://doi.org/10.1175/2009JCLI2719.1.
Welty, J., and X. Zeng, 2018: Does soil moisture affect warm season precipitation over the southern Great Plains? Geophys. Res. Lett., 45, 7866–7873, https://doi.org/10.1029/2018GL078598.
Wesely, M. L., 1976: The combined effect of temperature and humidity on the refractive index. J. Appl. Meteor., 15, 43–49, https://doi.org/10.1175/1520-0450(1976)015<0043:TCEOTA>2.0.CO;2.
Yang, L., G. Sun, L. Zhi, and J. Zhao, 2018a: Negative soil moisture-precipitation feedback in dry and wet regions. Sci. Rep., 8, 4026, https://doi.org/10.1038/s41598-018-22394-7.
Yang, L., and Coauthors, 2018b: A new generation of the United States National Land Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS J. Photogramm. Remote Sens., 146, 108–123, https://doi.org/10.1016/j.isprsjprs.2018.09.006.
Yin, J., J. D. Albertson, J. R. Rigby, and A. Porporato, 2015: Land and atmospheric controls on initiation and intensity of moist convection: CAPE dynamics and LCL crossings. Water Resour. Res., 51, 8476–8493, https://doi.org/10.1002/2015WR017286.
Zaitchik, B. F., J. A. Santanello, S. V. Kumar, and C. D. Peters-Lidard, 2013: Representation of soil moisture feedbacks during drought in NASA Unified WRF (NU-WRF). J. Hydrometeor., 14, 360–367, https://doi.org/10.1175/JHM-D-12-069.1.