**4. Results**

#### **4.1. Validation of simulated potential evapotranspiration and precipitation for present climate (1970: 2005)**

Analyzing together the multiannual averages obtained from observed data and those resulted from the five EURO-CORDEX models (**Table 1**), one can notice that, generally, the seasonal cycles of observed precipitation and potential evapotranspiration are captured by the ensemble of numerical experiments for the present climate (1970–2005). We illustrate here the results for the Prut basin (**Figures 4** and **5**) but the abovementioned conclusion stands for the other basins, too.

In general, the multimodel average simulates very well the observed potential evapotranspiration even though there are some underestimations, especially in spring months.

The results for model-simulated precipitation over the interval 1970–2005 do not reproduce the observed annual cycle as well as in the case of potential evapotranspiration. Even though the simulated annual cycle of precipitation generally resembles the observed one, the monthly values are generally overestimated. Underestimated monthly values are found in some summer months (July and August for the Prut basin).

was constrained by the data availability—they are mostly over the Romanian territory. In the case of Prut transboundary catchment, we used observations from two hydrometric stations: Dranceni in Romania and Brinza in Republic of Moldova. In general, there are quite large correlation coefficients illustrated in Figure 9 showing that the PDSI represents reasonably well

An interesting feature is the fact that ZIND correlations with Qmed are systematically larger than the correlation of P-PE with the Qmed, which implies that the Palmer model brings added value in assessing anomalies of the water deficit or surplus (ZIND) compared with the simple difference between precipitation and potential evapotranspiration (P-PE). Also, PRO correlations with Qmed are, in general, larger in cold season months compared to ZIND correlations with Qmed. This can be explained by the fact that the Palmer model does not take snow, frozen soil, and related processes into consideration—the precipitation is immediately transferred into the soil. That is why in winter months, any simultaneously precipitation-related correlations with Qmed are low. On the other hand, PRO depends on soil recharge linked to soil available capacity. Remarkably high correlations all over the year link basin-averaged PRO with Qmed at the Brinza station in the Prut catchment. An explanation could be that Brinza station is very close to the Prut outlet. Streamflows recorded at the Brinza station are integrating the runoff from the whole basin which is not the case for the streamflows recorded at the Dranceni station. However, the time interval used to compute these correlations at the Brinza station is shorter (1985–2015), implying lesser statistical significance. The results presented in **Figure 3** suggest that ZIND and PRO values could be used, at least for certain months and catchments, as simple and robust indicators of anoma-

lies in water cycle components (such as soil moisture and runoff) at the basin level.

**4.1. Validation of simulated potential evapotranspiration and precipitation for** 

ration even though there are some underestimations, especially in spring months.

Analyzing together the multiannual averages obtained from observed data and those resulted from the five EURO-CORDEX models (**Table 1**), one can notice that, generally, the seasonal cycles of observed precipitation and potential evapotranspiration are captured by the ensemble of numerical experiments for the present climate (1970–2005). We illustrate here the results for the Prut basin (**Figures 4** and **5**) but the abovementioned conclusion stands for the other basins, too. In general, the multimodel average simulates very well the observed potential evapotranspi-

The results for model-simulated precipitation over the interval 1970–2005 do not reproduce the observed annual cycle as well as in the case of potential evapotranspiration. Even though the simulated annual cycle of precipitation generally resembles the observed one, the monthly values are generally overestimated. Underestimated monthly values are found in some summer

the local process taking place in the analyzed catchments.

122 Engineering and Mathematical Topics in Rainfall

**4. Results**

**present climate (1970: 2005)**

months (July and August for the Prut basin).

**Figure 4.** Observed and simulated multiannual means (1970–2005) of monthly evapotranspiration (in mm/month) averaged over the Prut basin. The shaded band illustrates the simulated values from the five-member ensemble of regional climate experiments taken from the EURO-CORDEX archive (see **Table 1**). The black line represents the observation-derived values of potential evapotranspiration based on CRU data.

However, the fact that the multiannual pattern of the two water cycle components (precipitation and potential evapotranspiration) is reproduced, to some extent, by the regional climate models provides a certain level of confidence when analyzing their future evolution and related drought indices in the area of interests under climate change scenarios.

**Figure 5.** Observed and simulated multiannual means (1970–2005) of monthly precipitation (in mm/month) averaged over the Prut basin. The shaded band illustrates the simulated values from the five-member ensemble of regional climate experiments taken from the EURO-CORDEX archive (see **Table 1**). The black line represents the observation-derived values of potential evapotranspiration based on IMDROFLOOD gridded precipitation data.

#### **4.2. Future projections of water cycle components and climate change impact on drought at catchment level**

The main input data for computing the PDSI future projections are from the five-member EURO-CORDEX ensemble presented in **Table 1**.

**River basin Climate scenario Numerical experiments**

**1 2 3 4 5**

Variability and Change in Water Cycle at the Catchment Level

http://dx.doi.org/10.5772/intechopen.74047

125

Arges RCP 4.5 −0.75 −2.87 0.20 −2.28 −1.57

Mures RCP 4.5 1.85 −1.81 0.98 −0.94 −1.89

Prut RCP 4.5 −2.24 −1.10 −0.43 −1.81 −1.65

Siret RCP 4.5 −1.69 −1.18 −0.39 −2.75 −2.40

Somes RCP 4.5 2.12 −1.93 0.20 −0.75 −1.30

**Table 3.** PDSI change (in standardized units) for summer months (June–August) in the interval 1970–2100.

RCP 8.5 0.00 −3.81 −1.85 −4.01 −5.34

RCP 8.5 1.80 −0.31 −2.24 −1.22 −3.58

RCP 8.5 −1.14 0.39 −1.89 −2.83 −3.62

RCP 8.5 −0.79 −0.55 −1.85 −4.87 −5.27

RCP 8.5 0.83 −0.39 −2.95 0.02 −1.14

**Figure 6.** Simulated monthly values (1970–2100) of basin-averaged PDSI over the Prut catchment. The shaded band illustrates the simulated values from the five-member ensemble of regional climate experiments taken from the EURO-

CORDEX archive (see **Table 1**). The black line represents the multimodel ensemble mean.

The linear trend analysis of basin-averaged PDSI computed using the multimodel mean ensemble shows tendencies toward drought conditions over all basins, for both concentration scenarios, more pronounced in the summer months (**Table 2**). Also, as we expected, the trends are larger for higher concentration scenario (RCP 8.5). For instance, in summer months (June to August), the basin-averaged PDSI values are reduced in Arges and Siret basins with 2.98 and 2.67, respectively, over the interval 1970–2100 under RCP 8.5 (**Table 2**). A two-unit decrease in PDSI is consistent with the transition from normal conditions to moderate drought or from moderate to extreme drought. Pronounced decadal variability is also present in the future projections of PDSI under both RCP scenarios (e.g. **Figure 6**). In the Palmer classification [14], depending on the specific climate scenario and catchment, droughts that were deemed as incipient, mild or severe toward the end of the twentieth century will become a normal summer feature toward the end of the twenty-first century. The tendencies for meteorological droughts in summer are coming along with the downward trends of basin-averaged precipitation, potential runoff (and streamflow), and upward trend in potential evapotranspiration. For instance, the largest mean reduction in potential runoff in summer months is for Arges and Siret basins with 36 and 35%, respectively, over the interval 1970–2100 under RCP 8.5 (**Table 2**).


**Table 2.** PDSI mean change (in standardized units) and PRO mean change (in % relative to the mean of the interval 1970– 2005) for summer months (June–August) in the interval 1970–2100. Mean values of the five-member ensemble are used.


**4.2. Future projections of water cycle components and climate change impact on** 

The main input data for computing the PDSI future projections are from the five-member

The linear trend analysis of basin-averaged PDSI computed using the multimodel mean ensemble shows tendencies toward drought conditions over all basins, for both concentration scenarios, more pronounced in the summer months (**Table 2**). Also, as we expected, the trends are larger for higher concentration scenario (RCP 8.5). For instance, in summer months (June to August), the basin-averaged PDSI values are reduced in Arges and Siret basins with 2.98 and 2.67, respectively, over the interval 1970–2100 under RCP 8.5 (**Table 2**). A two-unit decrease in PDSI is consistent with the transition from normal conditions to moderate drought or from moderate to extreme drought. Pronounced decadal variability is also present in the future projections of PDSI under both RCP scenarios (e.g. **Figure 6**). In the Palmer classification [14], depending on the specific climate scenario and catchment, droughts that were deemed as incipient, mild or severe toward the end of the twentieth century will become a normal summer feature toward the end of the twenty-first century. The tendencies for meteorological droughts in summer are coming along with the downward trends of basin-averaged precipitation, potential runoff (and streamflow), and upward trend in potential evapotranspiration. For instance, the largest mean reduction in potential runoff in summer months is for Arges and Siret basins with 36 and 35%, respectively, over the interval

**drought at catchment level**

124 Engineering and Mathematical Topics in Rainfall

1970–2100 under RCP 8.5 (**Table 2**).

**River basin Climate scenario PDSI change in 131 years** 

Arges RCP 4.5 −1.45 −24

Mures RCP 4.5 −0.35 −12

Prut RCP 4.5 −1.45 −23

Siret RCP 4.5 −1.69 −22

Somes RCP 4.5 −0.35 −12

**(1970–2100)**

RCP 8.5 −2.98 −36

RCP 8.5 −1.10 −19

RCP 8.5 −1.81 −30

RCP 8.5 −2.67 −35

RCP 8.5 −0.75 −18

**Table 2.** PDSI mean change (in standardized units) and PRO mean change (in % relative to the mean of the interval 1970– 2005) for summer months (June–August) in the interval 1970–2100. Mean values of the five-member ensemble are used.

**PRO change in 131 years (1970–2100) (% of mean PRO computed from 1970 to 2005)**

EURO-CORDEX ensemble presented in **Table 1**.

**Table 3.** PDSI change (in standardized units) for summer months (June–August) in the interval 1970–2100.

**Figure 6.** Simulated monthly values (1970–2100) of basin-averaged PDSI over the Prut catchment. The shaded band illustrates the simulated values from the five-member ensemble of regional climate experiments taken from the EURO-CORDEX archive (see **Table 1**). The black line represents the multimodel ensemble mean.

However, when individual evolutions of PDSI values for each numerical experiment are analyzed, the intermodel and internal variability show up revealing cases in which the climaterelated signal of drought tendencies is not always present as for the ensemble mean. In **Table 3**, the PDSI changes for the interval 1970–2100 are presented for each numerical experiment listed in **Table 1** and for each river basin. This table illustrates the uncertainty associated with the signal revealed by the ensemble mean. The level of uncertainty related to the summer drought signal seems to be lowest for the Arges river basin if we count, from **Table 3**, the number of experiments for which there are higher PDSI changes in magnitude under the RCP 8.5 than those under the RCP 4.5 (four from five experiments). In this context, the level of uncertainty is highest for the Somes river basin (one from five experiments).

The results from the case studies based on the ensemble means of PDSI suggest that depending on the specific climate scenario and catchment, droughts that in the Palmer classification were deemed as incipient, mild or severe toward the end of the twentieth century, will become a normal summer feature toward the end of the twenty-first century in the mid and lower Danube basin. The tendency toward drought is present under the conditions of a reduction of runoff, mostly in summer, revealing the important role of potential evapotranspiration increase and precipitation decrease in the drought-related processes over the mid-latitude areas. However, the analysis of individual evolution of PDSI in the five numerical experiments, under both climate scenarios, reveals uncertainties associated with the identified signal of enhanced aridity at the basin level. The largest (lowest) uncertainty is found for the Somes (Arges) river basin. More studies that couple local climatic information with their hydrologic impact are needed to provide the background for the assessment of water resources under climate change conditions in terms of adaptation

Variability and Change in Water Cycle at the Catchment Level

http://dx.doi.org/10.5772/intechopen.74047

127

This chapter was supported by a grant of the Romanian National Authority for Scientific Research and Innovation, CCCDI—UEFISCDI, project number 81/2016 within PNCDI III"("Improving Drought and Flood Early Warning, Forecasting and Mitigation using real-time hydroclimatic indicators"—IMDROFLOOD) and by the European Commission financed under the ERA-NET Cofund WaterWorks2014 Call. This ERA-NET is an integral part of the 2015 Joint Activities developed by the Water Challenges for a Changing World Joint Programme Initiative (Water JPI). The description of Palmer Drought Severity Index presented in the first part of the Methodology section is taken from our work published in the reference [2]. We acknowledge the World Climate Research Programme's Working Group on Regional Climate, and the Working Group on Coupled Modeling, former coordinating body of CORDEX and responsible panel for CMIP5. We thank the climate modeling groups mentioned in **Table 1** for producing and making available their model output. Also, we acknowledge the Earth System Grid Federation infrastructure, an international effort led by the U.S. Department of Energy's Program for Climate Model Diagnosis and Intercomparison, the European Network for Earth System Modeling and the partners in the Global Organization

planning and sustainable development.

for Earth System Science Portals (GO-ESSP).

No conflict of interest is implied in this work.

**Conflict of interest**

**Acknowledgements**

Data are from multimodel means of a five-member ensemble consisting of five numerical experiments (see **Table 1**).

Data are from the five numerical experiments (see **Table 1**).
