Regional climate change projections for Europe agree in predicting a statistically significant warming in all seasons. The most significant climate change effect is its impact on water cycle through altering precipitation patterns and evapotranspiration processes at multiple scales. The anticipated changes in the distribution and precipitation amounts together with continuously increasing temperatures may induce a higher rate of water consumption in plants, which can generate changes in soil moisture, groundwater, and the water cycle. Thus, climate change can cause changes in the water balance equations structure. A Thornthwaite-type monthly step water balance model was established to compare the water balance in three different surface land cover types: (i) a natural forested area; (ii) a parcel with mixed surface cover; (iii) an agricultural area. The key parameter of the model is the water storage capacity of the soil. Maximal rooting depth of the given area is also determinable during the calibration process using actual evapotranspiration (AET) and soil physical data. The locally calibrated model was employed for assessing future AET and soil moisture of selected land cover types using data from four bias-corrected regional climate models. The projections demonstrate increasing actual evapotranspiration values in each surface cover type at the end of the 21st century. Regarding the 10th percentile minimum soil moisture values, the forested area displayed an increasing trend, while the agricultural field and mixed parcel showed a strong decrease. The 30-year monthly means of evapotranspiration shows the maximum values in June and July, while the minimum soil moisture in September. Water stress analysis indicates water stress is expected to occur only in the agricultural field during the 21st century. The comparison of the three surface covers reveals that forest has the greatest soil water storage capacity due to the highest rooting depth. Thus, according to the projections for 21st century, less water stress is predicted to occur at the forested area compared to the other two surface covers which shows shallow rooting depth.
The ongoing climate change can be characterized by a statistically significant warming trend in all seasons throughout Europe (
Higher temperatures reflect larger energy potentials in the atmosphere, which, therefore, contains more water at the same time and/or has a shorter water vapour retention time, both of which will accelerate the hydrological cycle (
The most significant effect of climate change is its impact on the water cycle by altering precipitation patterns and evapotranspiration processes at multiple scales (
In the Carpathian Basin, 90% of the precipitation is evapotranspired, while the remaining 10% is runoff (
Previous studies modeling the impact of climate change on water balance (
The overall objectives of this study are to establish a monthly step water balance model for three different surface covers: (i) a forested area; (ii) a mixed parcel (1 km2 corn field plot with strips of poplar trees); (iii) an agricultural area. Using the calibrated model, the impact of climate change in the 21st century is analysed to compare the water balance of forest and two other land use types located at the north-western part of the Carpathian Basin.
To test our model, we used three study areas in the Carpathian Basin, namely: a forested area, a mixed parcel, and an agricultural field. The first two are situated in the Western part of the Transdanubian Region of Hungary, while the third is in the eastern part of Austria next to Vienna (see Fig. S1 in Supplementary material).
The forested area is an experimental catchment at the eastern foothills of the Alps near the city of Sopron. The elevation of the study area ranges from 370 to 550 m a.s.l.
This area has a subalpine climate, with an average annual temperature of 8.5 °C, and annual precipitation of 700-750 mm. The driest season is autumn, while the wettest is late spring and early summer (
The geological basis of the catchment is fluvial sediments deposited in five distinct layers in the tertiary (Miocene) period on crystalline bedrock. A finer-grained layer appears in the valley bottom, which is a good aquifer, giving rise to perennial streams (
The investigation period was from January 2000 to December 2008, due to the availability of the data which originated from the AgroClimate.2 project (
Mean monthly temperature (TM) and monthly summed precipitation (PM) values are utilized as inputs for the forested area and measured (remote-sensing based) actual evapotranspiration values (ETCREMAP) for calibration and validation.
The mixed parcel is basically used as an agricultural plot and was formerly a cornfield, except during the period 2003-2007 when it was used to grow barley and in 2004 when it was used for wheat. However, poplar (
The selected parcel is located in the Mosoni-sík microregion that is situated in Györ-Moson-Sopron County. It is basically (73.5%) plough-land. On the whole, this natural microregion is an alluvial plain (
The climate is continental, with temperature differences between the western and eastern part of the natural microregion. The average annual temperature is 9.7 °C and the annual precipitation is 560 mm (
The investigation period was from January 2000 to December 2008, the same as for the forested area. The data originated from the AgroClimate.2 project (
The input dataset contains mean monthly temperature (TM) and monthly summed precipitation (PM) values as well as measured (remote-sensing) actual evapotranspiration values (ETCREMAP) for calibration and validation.
The agricultural field is an area of about 1000 km2 in the eastern part of Austria, between Vienna and the border of Slovakia. The elevation of this study area is 157 m a.s.l. It is characterized by a subhumid climate with a mean annual temperature and precipitation of approximately 10 °C and 550 mm, respectively. Typical summers are hot and dry, while winters are mainly cold with severe frost and limited snow cover (
Basic data for this study were obtained at the experimental farm of the University of Natural Resources and Life Sciences, Vienna (BOKU), in Groß-Enzersdorf (48° 12′ N, 16° 34′ E; 157 m a.s.l.). Regarding climate conditions, the location is representative for the agricultural field (
Mass changes of the surface-soil-vegetation system and, therefore, changes in soil water content can be estimated using weighing lysimeters (
These soil characteristics and the consequent hydraulic properties were taken as the basis for the simulation. Although the input parameters are assumed to be typical for the region, a single soil profile cannot represent the characteristics of a large area like the agricultural field in general.
The utilized lysimeter and the surrounding area were permanently covered by grass and maintained to represent the standard conditions for the determination of reference evapotranspiration (
Therein, ΔWlys is the daily change of the soil water mass, determined by means of the weighing facility. ΔWdrain is the daily amount of drainage water (freely draining by gravity forces), measured at the bottom outlet by means of a tipping bucket device. Plys is precipitation and Ilys is irrigation on the lysimeter. Both were assumed to occur when net changes in soil water (ΔWlys + ΔWdrain) were positive. Consequently, evapotranspiration from the lysimeter surface (ETlys) was quantified as negative changes in soil water (ΔWlys + ΔWdrain). All units are millimetres per day (mm day-1). Specific smoothing functions were applied to improve measurement accuracy, which is 0.1 mm at average wind velocity. A detailed description of measurements and data processing can be found in Nolz et al. (
Mean monthly temperature (TM) and summed monthly precipitation (PM) values make the input dataset, and the actual evapotranspiration values (ETLYSIMETER) are for calibration and validation. Nevertheless, all input data refer to the experimental site in Groβ-Enzersdorf. In the given case, irrigation and precipitation comprise the input parameter.
The investigation period was from January 2004 to December 2011. It is important to note that the difference between the time series of the study areas is due to the availability of the input data.
The Thornthwaite-type water balance model represents a 1-D system which considers only vertical fluxes. Input values are monthly precipitation (PM, mm) and temperature (TM, °C -
The first step in setting up the model was the calculation of the potential evapotranspiration (PET). PET is the amount of water that can be evaporated and transpired when soil water is sufficient to meet atmospheric demand (
where D is the day length (hr), TM is the average monthly temperature (°C), and e*m is the saturation vapour pressure (kPa).
The next step was a condition. If PM ≥ PETM then ETM = PETM and (
where PETM is the calibrated monthly potential evapotranspiration (mm). Determination of PETM is part of the calibration, which will be described later. ETM (mm month-1) is the monthly actual evapotranspiration, and SOILM (mm) is the monthly soil moisture. Actual evapotranspiration is the amount of water evaporating from the surface and transpired by plants if the total amount of water is limited (
For the simulation procedure, the first SOILM-1 value was set to a maximum value that corresponds with the soil-water storage capacity (SOILMAX, mm). The basic assumption was that soil is saturated before the beginning of the vegetative period. SOILMAX was introduced using unsaturated hydraulic parameters of the study areas’ soil types and setting a rooting depth of 1 m (
where
The following procedure illustrates how soil water storage is considered as a reservoir for evapotranspiration: if precipitation is less than the (calibrated) potential evapotranspiration in a certain month (
where (
and ΔSOIL is the decrease in soil water storage [mm].
Remote sensing based (for the forested area and mixed parcel) and grass-covered lysimeters (for the agricultural field) actual evapotranspiration data served as basis for calibration and validation. The available time series for the forested area and for the mixed parcel (2000-2008) was divided into two parts. The first part is used for calibration from 2000 to 2005, whereas the second is for validation from 2006 to 2008. In the agricultural field, the time series (2004-2011) was also divided into two parts. The first (from 2004 to 2008) was used for calibration and the second (from 2009 to 2011) for validation.
The calibration datasets were further divided into two parts considering both potential and actual evapotranspiration. The results of calibration and validation are reported in the Results section.
Fig. S2 (Supplementary material) schematically represents the functioning of the model and the relationships between the applied parameters in the modelling process for the forested area and mixed parcel.
Parameters of the calibration and the input data (TM and PM) of the validation period (2009-2011 for the agricultural field; 2006-2008 for the forested area and the mixed parcel) were used for the validation.
As the basis for the projection procedure, the water balance model was re-calibrated for each study area using all available data (2000-2008 for the forested area and mixed parcel; and 2004-2011 for the agricultural field). This was done because model calibration using as much data as possible was assumed to deliver the best possible results.
Inputs for predicting future developments of actual evapotranspiration (ETM), soil moisture (SOILM), and the lower 10th percentile of soil moisture (SOILM_10Perc,
FORESEE is a bias-adjusted database that contains daily meteorological data (min/max temperature and precipitation) based on the simulation results of ten RCMs for 2015-2100, and observation based data for the period 1951-2014 interpolated to 1/6×1/6 degree spatial (horizontal) resolution grid (using inverse distance interpolation technique). Furthermore, all of the time series were converted to a 365-day calendar (
The four different RCMs illustrate the uncertainties, because all climate projections have inherent uncertainties. Data were extracted from the pixel including the study sites’ coordinates. The main properties of the RCMs can be found in Tab. S1 (Supplementary material). In the following, each model will be referred to by their model ID (first column of Tab. S1).
The time scale of RCMs covers a range from 2015 to 2100. Each contains temperature and precipitation data in monthly time intervals. To evaluate the results for the 21st century, four main investigation periods were designated: 1985-2015 (01.01.1985 - 01.01.2015), 2015-2045 (01.01.2015 - 01.01.2045), 2045-2075 (01.01.2045 - 01.01.2075), and 2070-2100 (01.01.2070 - 01.01.2100). The results of the first investigation period (1985-2015) are based on observation data (model ID “0”). As mentioned before, the FORESEE results for the RCMs were available from 2015; therefore, the investigation periods had to be shifted by five years compared to the investigation periods of the AgroClimate.2 project. With the data at hand, these 30-year-blocks with a five-year overlap in the last two periods seemed the best partitioning. The overlap in the last part of the 21st century was necessary because only 25 years of data were available.
An appropriate and simple way to assess water stress is the calculation of the relative extractable water (REW, dimensionless) using the following equation (
When REW drops below 50% of SOILMAX, the transpiration is progressively reduced (because of stomatal closure); hence, plant water stress is assumed to occur. SOILMAX parameter is the maximal amount of water available to plants and, therefore, it reflects the maximum extractable water in the soil. The average soil moisture (SOILM) is the extractable water in the different periods of investigation.
Model performance was tested using the coefficient of determination (R2) and the Nash-Sutcliffe model efficiency coefficient (R2NS). The latter is a criterion used for calibration and validation of hydrologic models (
where ETMSR_i is the time series of measured values, ETSIM_i is the time series of simulated values, and mMSR_i is the average value for the considered period.
The first step of calibration considered the potential evapotranspiration for actual land cover using ETCREMAP-values (for the forested area and mixed parcel) and Etlys values (for the agricultural field) at well-watered conditions. The latter were assumed to occur when precipitation or the actual evapotranspiration (ETCREMAP or ETlys) exceeded the potential evapotranspiration (PETH -
The ETlys/ETCREMAP values selected in such a way are denoted PETlys/PETCREMAP. Measured (PETlys/PETCREMAP) and calculated (PETH) values were correlated with the second variable as the explanatory one. PET is known to be different between growing season and dormancy, therefore different relationships had to be established for the two periods (
We compared the three study areas regarding PET calibration. Correlation between PETH and PETCREMAP/LYS during the period of dormancy is illustrated by the section on the left of the vertical dotted line (broken-line approach) in
The 1:1 dotted lines exposed overestimation in the forested area (
The breakpoint value obtained for the forested area (24.3 mm) is lower than the two other areas (mixed parcel: 39.1 mm; agricultural field: 36.9 mm). This can be attributed to the presence in the forested area of conifer species, the growing season of which begins earlier. Nevertheless, the value of albedo is also smaller in the case of the forested area; consequently the absorbed energy is higher, which can be manifested in higher evapotranspiration.
By contrast, each study area expresses greater or lesser underestimation in the growing season (
As the second step of the calibration, the calculated actual evapotranspiration (ETM) has been calibrated based on the parameter SOILMAX. In this case, the initial estimate of SOILMAX had to be adjusted in order to reach a maximal correlation between ETlys/ETCREMAP and ETM. To achieve this maximum correlation, the “optim” function of the mentioned R software was applied. Using the value of SOILMAX after calibration, the vertical extent of the root zone (and the maximum depth of tilth) can be calculated using soil texture data.
The Nash-Sutcliffe coefficient (R2NS) of the calibrated models were 0.85, 0.88 and 0.88 for the forested area, mixed parcel and agricultural field, respectively, while the coefficients of determination R2 were 0.88, 0.86, 0.89, respectively (
Greater difference was found between the measured ETCREMAP and the calculated ETM values in the forested area, particularly in the summer of 2007 (
Although the curves of the agricultural field model fit each other the best visually, this model performed the “worst” regarding the Nash-Sutcliffe coefficient, due to the data loss caused by a thunderstorm in the summer of 2009.
As mentioned above, the model was re-calibrated for each study area using all available data as a basis for the projection procedure. The parameters of the re-calibrated models used in the projection phase are reported in
We used another method to determine the rooting depth at the agricultural field because
The annual temperature means and the annual precipitation sums show an increasing tendency for each study site towards the end of 21st century. According to the RCM projections, the rate of increase of annual temperature in the period 2070/ 2100 (compared to the 1985/2015 reference period) is 1.9 °C for all the studied areas, while for precipitation the projections are: 68 mm (forested area); 69 mm (mixed parcel); 71 mm (agricultural field). Therefore, the rates of the expected temperature and precipitation increase are equivalent for the three study areas.
The different RCMs used for our projection provide different results, which influence the parameters (outputs) of the water balance. Compared to the averages of RCMs, the model with higher precipitation may indicate higher available water, while the greater temperature may cause greater potential evapotranspiration.
Contrary to the tendencies of ETM values, there are larger differences in the mean values of soil moisture (SOILM) among the study sites (
With regard to plant water uptake, the minimum soil moisture were calculated as 10th percentile minimum values (SOILM_10Perc -
The previous analyses are based on annual mean values for the four 30-year-long investigation periods. These analyses, however, do not indicate the monthly development of output parameters of the models. Hence, a future research study that focuses on the 30-year monthly mean of ETM plus SOILM is required.
The 30-year monthly mean of SOILM shows a slight increase from January to March, when the soil is saturated and the values of SOILM are the closest to the water storage capacity (SOILMAX). From March to September, a decrease of soil moisture occurs due to the rising evapotranspiration. Consequently, the minimum values occur in early autumn (September) in each study area (
When the three study sites are compared, the most significant differences appear for the 30-year monthly mean of SOILM values rather than at the ETM values. The highest SOILM and ETM values are observed at the forested area. The forested area and mixed parcel show equal annual fluctuation (~150 mm), whereas the lowest values of ETM and SOILM and smallest fluctuation of SOILM (~90 mm) are found at the agricultural field. The rates of the annual soil moisture fluctuations and soil moisture storage capacity (SOILMAX) are lowest at the forested area (30%) but highest at the agricultural field (63%).
Based on the above evidence, the water stress probability may increase towards the end of the 21st century.
Based on the above evidences, significant water stress is expected in the future at the agricultural field area, due to the relatively small SOILMAX value and small rooting depth detected.
In this study, a Thornthwaite-type water balance model was adapted and applied to assess the future development of evapotranspiration and soil moisture in the western part of the Carpathian Basin.
Our study indicates an increasing tendency of actual evapotranspiration across the study areas towards the end of the 21st century, with high annual fluctuation and greater peaks for summer. The monthly average values of soil moisture, however, show no clear trend or a weak increase, whereas the lower 10th percentile minimums show a significant decrease and greater annual fluctuation (particularly in the early autumn) towards the end of the 21st century. The results showed that significant plant water stress is expected to occur only at the agricultural field.
In this study a relatively straightforward model approach was applied to regional conditions, though further research should refine our analysis, for example by considering crop characteristics, different soil, or land use changes. According to
The monthly average of the modeled relative extractable water (REW) for the crop field showed an increasing trend (from 42% to 46%) towards the end of the 21st century, and values below the 50% threshold are expected to occur frequently. In contrast, REW values for forested area and mixed parcel do not approach such threshold (78% and 71%, respectively). According to
Regarding the annual tendency of AET rates, our results agree with those reported by
Climate simulations for the Rhine-Meuse drainage area in central-western Europe were carried out by
In this study, a Thornthwaite-type water balance model was applied to assess the impact of climate change on water balance components in a natural forested area, a mixed parcel, and an agricultural field in the western Carpathian Basin.
The main advantage of the developed model is the low amount of input data required (temperature and precipitation). This allows the model to be easily applied to other places where measured data are available for calibration/validation. This model system ensures fast impact analysis of climate change on evapotranspiration and soil water storage, along with the calculation of water stress parameters. Moreover, the model requires a significantly lower amount of work for input data pre-processing and baseline investigations than more complex models.
Based on standard future climate scenarios, an increase in ETM is expected in future decades, with a remarkable shift predicted for SOILM, indicating that less soil water will be available for plant growth during summer months in the studied region, and confirming the evidences reported by studies using similar methods (
AET: actual evapotranspiration (mm); CREMAP: Calibration-Free Evapotranspiration Mapping; D: daylength (hour); e*m:saturation vapor pressure (kPa); ETCREMAP: remote-sensing based actual evapotranspiration (mm); ETLYSIMETER: actual evapotranspiration values measured by weighing-lysimeter (mm); ETM: monthly actual evapotranspiration (mm · month-1): GCM: general circulation model; PAW: plant available water (mm); PET: potential evapotranspiration (mm); PETCREMAP: remote-sensing PET based on actual evapotranspiration at well-watered conditions (mm); PETLYSIMETER: actual evapotranspiration values measured by weighing-lysimeter at well-watered conditions (mm); PETH: Hamon type potential evapotranspiration (mm); PETM: calibrated monthly potential evapotranspiration (mm); PM: monthly summed precipitation (mm); R2: coefficient of determination (dimensionless); RCM: regional climate model; REW: Relative Extractable Water (dimensionless); R2NS: Nash-Sutcliffe coefficients (dimensionless); SOILM: monthly soil moisture (mm); SOILMAX: soil-water storage capacity (mm); SOILM_10Perc: 10th percentile soil moisture minimum values (mm); TM: monthly summed temperature (°C); zrz: rooting depth (vertical extent of root zone, mm); ΔWlys: daily change of soil water mass (mm day-1); Δwdrain: daily change of drainage water (mm day-1); Δplys: daily change of precipitation (mm day-1); ΔIlys: daily change of irrigation (mm day-1); ΔETlys: daily change of evapotranspiration (mm day-1): ΔSOIL: decrease in soil storage (mm); θfc: water content at field capacity (dimensionless); θpwp: water content at permanent wilting point (dimensionless).
This work was supported by the project EFOP-3.6.2-16-2017-00018 of the University of Sopron, Hungary. Fundings from the János Bolyai Scholarship of the Hungarian Academy of Sciences and from the Ministry of Agriculture in Hungary are gratefully acknowledged. The authors wish to thank Reinhard Nolz (University of Natural Resources and Life Sciences, Vienna, Austria) for supplying the lysimeter data.
Relationship between PETCREMAP/PETLYSIMETER and PETH in growing and dormant seasons with a 1:1 line (dotted), at forested area (a), at mixed parcel (b), at agricultural field (c) obtained after the calibration of PETH. The red triangles represent the values of the dormancy period, while blue dots represent the values of the growing season. The vertical dotted line is the separation of the two different vegetative states.
Relationship between the calculated ETM and the measured ETCREMAP or ETLYSIMETER obtained after model calibration in each study area. (a): forested area; (b): mixed parcel; (c): agricultural field.
Comparison of the time series of measured ETCREMAP or ETLYS and calculated ETM values obtained after the validation step in each study area. (a): forested area; (b): mixed parcel; (c): agricultural field.
Monthly values of ETM for the study areas for the investigated 30-year means. (a): forested area; (b): mixed parcel; (c): agricultural field.
Monthly values of SOILM for the study areas for the investigated 30-year means. (a): forested area; (b): mixed parcel; (c): agricultural field.
Results of the adjusted, re-calibrated model parameters in the study sites.
Study sites | Re-calibrated PET parameter | Re-calibrated AET parameter | |||
---|---|---|---|---|---|
Model | R2 | Model | R2 | R2NS | |
Forested area | PETM = 0.42 · PETH + 1.09 · (PETH - 26.04) | 0.98 | ETCREMAP = 1.14 · ETM - 4.79 | 0.89 | 0.88 |
Mixed parcel | PETM = 0.50 · PETH + 1.05 · (PETH - 37.13) | 0.98 | ETCREMAP = 1.08 · ETM - 4.31 | 0.87 | 0.88 |
Agricultural field | PETM = 0.54 · PETH + 1.04 · (PETH - 36.79) | 0.98 | ETLYS = 1.04 · ETM - 2.36 | 0.88 | 0.88 |
Soil types, values of field capacity, permanent wilting point, re-calibrated SOILMAX and re-calibrated rooting depth in the study areas. Soil types were determined using the available data in the AgroClimate.2 project (forested area) or by soil sampling from borehole (mixed parcel). Field capacity (FC, dimensionless) and permanent wilting point (PWP, dimensionless) values of forested area and mixed parcel were used in accordance with
Study sites | Soil type | FC | PWP | SOILMAX(mm) | Rooting depth (mm) |
---|---|---|---|---|---|
Forested area | sandy loam | 0.207 | 0.095 | 502.4 | 4486 |
Mixed parcel | sandy loam | 276.9 | 2472 | ||
Agricultural field | sandy loam | * | * | 142.4 | 890 |
Main properties of the soil profile in the lysimeter. (PAW): plant available water; (PAWacc): PAW accumulated to the bottom of the given layer.
Depth(cm) | PAW(vol-%) | PAWacc(mm) | ||
---|---|---|---|---|
0-20 | 30.1 | 14.9 | 15.2 | 30.4 |
20-40 | 32.7 | 17.2 | 15.5 | 61.4 |
40-60 | 30.4 | 14.7 | 15.7 | 92.8 |
60-80 | 30.2 | 13.5 | 16.7 | 126.2 |
80-100 | 29.7 | 12.3 | 17.4 | 161 |
100-140 | 30.0 | 11.9 | 18.1 | 233.4 |
140-250 | 1.7 | 0.8 | 0.9 | - |
ETM, SOILM and SOILM_10Perc values (± standard deviation) obtained from the projection at the study areas.
Study sites | Parameters | 1985/2015 | 2015/2045 | 2045/2075 | 2070/2100 |
---|---|---|---|---|---|
Forested area | ETM (mm month-1) | 48 ± 38 | 48 ± 37 | 51 ± 39 | 52 ± 40 |
SOILM (mm) | 417 ± 92 | 416 ± 74 | 415 ± 76 | 394 ± 86 | |
SOILM_10Perc (mm) | 208 ± 59 | 270 ± 32 | 271 ± 25 | 234 ± 37 | |
Mixed parcel | ETM (mm month-1) | 43 ± 35 | 43 ± 33 | 45 ± 35 | 46 ± 35 |
SOILM (mm) | 215 ± 57 | 210 ± 61 | 211 ± 63 | 199 ± 69 | |
SOILM_10Perc (mm) | 109 ± 20 | 96 ± 15 | 96 ± 14 | 77 ± 21 | |
Agricultural field | ETM (mm month-1) | 49 ± 34 | 49 ± 33 | 52 ± 34 | 53 ± 35 |
SOILM (mm) | 58 ± 40 | 65 ± 43 | 66 ± 44 | 67 ± 48 | |
SOILM_10Perc (mm) | 8 ± 3 | 7 ± 2 | 6 ± 3 | 5 ± 3 |
The values of relative extractable water (REW) in the 21st century for the study areas. Each 30-year period contains 360 months. The value 0.83 indicates that 83% of the 360 monthly REW values do not drop under the threshold of 50% SOILMAX.
Study area | 1985/2015 | 2015/45 | 2045/75 | 2070/2100 |
---|---|---|---|---|
Forested area | 0.83 | 0.83 | 0.82 | 0.78 |
Mixed parcel | 0.78 | 0.76 | 0.76 | 0.71 |
Agricultural field | 0.42 | 0.46 | 0.46 | 0.46 |
Tab. S1 - The main properties of the applied RCMs.
Fig. S1 - The location of study areas.
Fig. S2 - Graphical representation of the model of the study areas.