**4. Temporal evolution of drought-related variables**

Finally, we investigated the relationship between the SPEI and other droughtrelated variables. **Figure 6** shows the temporal evolution of the standardized anomalies of SFCEVP accumulated at 3 months as well as of the monthly mean SM for all river basins. The SPEI-03 was also displayed with comparative purposes. In broad terms, the SM evolves similar to the SPEI-03, showing correlations above 0.7 in all river basins (second column in **Table 3**). However, the SFCEVP presents more

**139**

**Figure 6.**

*Understanding the Drought Phenomenon in the Iberian Peninsula*

*Standardized anomalies of the 3 months accumulated SM (blue line) and SFCEVP (orange line), SPEI-03 (gray area) evolutions for the NA, MS, DU, EB, NE, PB, TJ, SE, GU, GQ, SB, and BI river basins.*

**River basins SPEI-SM SPEI-SFCEVP SPEI-runoff** NA 0.85\* 0.24 0.91\* MS 0.84\* 0.49\* 0.91\* DU 0.82\* 0.56\* 0.80\* EB 0.81\* 0.53\* 0.80\* NE 0.81\* 0.67\* 0.80\* PB 0.84\* 0.60\* 0.87\*

*DOI: http://dx.doi.org/10.5772/intechopen.85472*

#### **Figure 6.**

*Drought - Detection and Solutions*

map), the positive trends are less frequent, showing an increased trend in more than one parameter only for the PB, the DU, and the EB river basins. In these three basins, the results showed an increase in both severity and duration of drought events.

Finally, we investigated the relationship between the SPEI and other droughtrelated variables. **Figure 6** shows the temporal evolution of the standardized anomalies of SFCEVP accumulated at 3 months as well as of the monthly mean SM for all river basins. The SPEI-03 was also displayed with comparative purposes. In broad terms, the SM evolves similar to the SPEI-03, showing correlations above 0.7 in all river basins (second column in **Table 3**). However, the SFCEVP presents more

**4. Temporal evolution of drought-related variables**

**138**

*Standardized anomalies of the 3 months accumulated SM (blue line) and SFCEVP (orange line), SPEI-03 (gray area) evolutions for the NA, MS, DU, EB, NE, PB, TJ, SE, GU, GQ, SB, and BI river basins.*


#### *Drought - Detection and Solutions*


#### **Table 3.**

*Pearson's temporal correlation (r) between the SPEI-03 and the soil moisture content (SPEI-SM) and SFCEVP (SPEI-SFCEVP), and between the SPEI-12 and the 12 months' accumulated runoff (SPEI-runoff).*

**141**

**Figure 7.**

*Understanding the Drought Phenomenon in the Iberian Peninsula*

discrepancies, especially for the NA and the MS river basins. This latter is probably produced because wet zones such as the northwestern IP are characterized by a SFCEVP less sensitive to changes in SM, being energy-limited regions. Here, only under very long periods of precipitation scarcity, the SFCEVP is strongly affected by the water availability. Contrariwise, in transitional zones (i.e., southern IP), the SFCEVP is usually limited, and then, the SM, and consequently the precipitation, largely controls the SFCEVP. In this region (e.g., the GQ or the SE river basins), the SFCEVP agrees really well with the SPEI-03. This behavior is reflected in the temporal correlations between the SFCEVP and the SPEI-03 (third column in **Table 3**), which tends to be higher over the southern river basins. In broad terms, the Pearson's

*Standardized anomalies of the 12 months accumulated runoff (blue line) and the SPEI-12 (gray area)* 

*temporal series for NA, MS, DU, EB, NE, PB, TJ, SE, GU, GQ, SB, and BI river basins.*

*DOI: http://dx.doi.org/10.5772/intechopen.85472*

**Figure 7.**

*Drought - Detection and Solutions*

*Significant correlations at 95% confidence level.*

*\**

**Table 3.**

**River basins SPEI-SM SPEI-SFCEVP SPEI-runoff** TJ 0.79\* 0.56\* 0.84\* SE 0.80\* 0.70\* 0.80\* GU 0.76\* 0.60\* 0.80\* GQ 0.77\* 0.61\* 0.81\* SB 0.75\* 0.67\* 0.81\* BI 0.76\* 0.63\* 0.77\*

*Pearson's temporal correlation (r) between the SPEI-03 and the soil moisture content (SPEI-SM) and SFCEVP* 

*(SPEI-SFCEVP), and between the SPEI-12 and the 12 months' accumulated runoff (SPEI-runoff).*

**140**

*Standardized anomalies of the 12 months accumulated runoff (blue line) and the SPEI-12 (gray area) temporal series for NA, MS, DU, EB, NE, PB, TJ, SE, GU, GQ, SB, and BI river basins.*

discrepancies, especially for the NA and the MS river basins. This latter is probably produced because wet zones such as the northwestern IP are characterized by a SFCEVP less sensitive to changes in SM, being energy-limited regions. Here, only under very long periods of precipitation scarcity, the SFCEVP is strongly affected by the water availability. Contrariwise, in transitional zones (i.e., southern IP), the SFCEVP is usually limited, and then, the SM, and consequently the precipitation, largely controls the SFCEVP. In this region (e.g., the GQ or the SE river basins), the SFCEVP agrees really well with the SPEI-03. This behavior is reflected in the temporal correlations between the SFCEVP and the SPEI-03 (third column in **Table 3**), which tends to be higher over the southern river basins. In broad terms, the Pearson's correlations also show that the SPEI-03 correlates well in general with the SFCEVP, showing significant correlations at the 95% confidence level for all river basins, except over the NA river basins.

The hydrological droughts are represented in **Figure 7**. Here, the temporal evolution of the runoff, which was obtained by aggregated standardized anomalies at 12 months, was displayed, as well as the SPEI-12. As for the previous analysis, different periods with positive and negative anomalies were recorded in all river basins, reproducing wet and dry periods similar to the SPEI-12. In fact, the SPEI-12 reflects long-term precipitation patterns, being thus useful to determine streamflow and reservoir levels. For instance, a marked wet period clearly appeared during 2001 over the NA, the MS, the DU, the PB, and the TJ river basins in both the runoff and the SPEI-12 temporal series. In the same way, certain river basins presented marked negative standardized anomalies of runoff during 2005, which also showed important negative values of SPEI-12. As expected, r between the SPEI-12 and the runoff was really high showing values around 0.85, with the highest ones over the NA and the MS river basins (r values of 0.91).

## **5. Conclusions**

This chapter is devoted to analyzing spatiotemporal patterns of droughts over the IP, a region particularly vulnerable to this extreme event [7] for the period 1980–2014. To do this, a regional climate simulation using the WRF model was completed, obtaining thus high spatiotemporal resolution climate data. Temperature and precipitation data from WRF were used to compute the SPEI at 3 and 12-month time scales with the purpose of analyzing different drought applications (e.g., agricultural and hydrological). This index has proved to be adequate to analyze droughts in the context of global warming.

Firstly, the regional performance of the SPEI was assessed. The most general evolution in droughts is consistent with other studies that allocated the main drought episodes occurred over the IP (e.g., [27]). The results evidenced that WRF appears as a promising tool for analyzing droughts since they provide a great number of climate variables at adequate spatial and temporal resolutions, which are unusually obtained from observations. In fact, climate data from regional simulations may be useful to determine where and when drought is occurring, providing valuable information to compute different drought indices using different droughtrelated variables.

These results also show that the SPEI-03 represents more frequent and shorter episodes, with the median duration and intensity of about 3–4 months and 1, respectively. In this regard, the largest events occurred over the southern IP, which is especially vulnerable due to its aridity conditions. In fact, this region is arid and then major damages may be produced under the occurrence of severe drought events.

On the other hand, the SPEI-12 presented longer drought episodes, showing, in general, more spread in its duration distributions, with the median values ranging from 5 to 17 months. However, in regard to the intensity, droughts presented similar behavior for both time scales.

Concerning the existence of the trends in droughts along the analyzed period, the results indicate that, in general, an increase in at least one of the drought properties occurred over the period 2000–2014 with respect to the period 1980–1999 in most of the river basins, more evident for the SPEI-03.

In the comparison between the SM and the SPEI-03, the results show that this variable presents a good agreement with the SPEI, which is an index based on a balance between water supply and potential demand at 3 months, which suggests

**143**

the future.

from NCAR.

**Acknowledgements**

**Conflict of interest**

conflict of interest.

**Author details**

provided the original work is properly cited.

Yolanda Castro-Díez and María Jesús Esteban Parra

\*Address all correspondence to: mgvaldecasas@ugr.es

*Understanding the Drought Phenomenon in the Iberian Peninsula*

SFCEVP is not a limiting factor such as the northwestern IP.

that the 3-month time scale is adequate to study agricultural drought in general over the IP. In fact, it is well known that variations in the SM are the result of changes in precipitation, SFCEVP, and runoff. Therefore, it is an accumulated measure of water availability, and thus, it can be used to detect dry and wet periods [32]. On the other hand, the SFCEVP also affects the agricultural systems, but its relationship with droughts is more complex. That is, the SFCEVP is the actual loss of water, which depends on the water demand by plants and the atmosphere (i.e., the potential evapotranspiration and consequently the temperature), but also on the water availability. This is clearly shown in arid regions, in which the SM largely controls the SFCEVP. Therefore, the actual water balance can be used as a proxy of drought occurrences, but it must be interpreted with caution in regions in which the

Hydrological droughts can be defined as the scarcity of surface and subsurface water resources that leads to negative effects in water resources management system. In this regard, our results indicate that the annual runoff seems to be a good variable to identify hydrological droughts, showing an especial good agreement with the SPEI-12. The agreement between variables shown here also evidences that the SPEI is an adequate index to investigate drought conditions over the IP, reproducing these in a similar way than other drought-related variables over the study period. This is of high relevance to adequately develop strategies to mitigate the effects of droughts in

This study has been financed by the Spanish Ministry of Economy and Competitiveness, with additional support from the European Community fund (FEDER), project CGL2013-48539-R and CGL2017-89836-R. The WRF simulation was run in the ALHAMBRA supercomputer infrastructure (https://alhambra.ugr. es). The ERA-Interim data were obtained from ECMWF portal and the WRF model

*DOI: http://dx.doi.org/10.5772/intechopen.85472*

© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,

Matilde García-Valdecasas Ojeda\*, Emilio Romero Jiménez, Sonia R. Gámiz-Fortis,

Department of Applied Physics, University of Granada, Granada, Spain

The authors declare that the research was conducted in absence of any potential

#### *Understanding the Drought Phenomenon in the Iberian Peninsula DOI: http://dx.doi.org/10.5772/intechopen.85472*

that the 3-month time scale is adequate to study agricultural drought in general over the IP. In fact, it is well known that variations in the SM are the result of changes in precipitation, SFCEVP, and runoff. Therefore, it is an accumulated measure of water availability, and thus, it can be used to detect dry and wet periods [32]. On the other hand, the SFCEVP also affects the agricultural systems, but its relationship with droughts is more complex. That is, the SFCEVP is the actual loss of water, which depends on the water demand by plants and the atmosphere (i.e., the potential evapotranspiration and consequently the temperature), but also on the water availability. This is clearly shown in arid regions, in which the SM largely controls the SFCEVP. Therefore, the actual water balance can be used as a proxy of drought occurrences, but it must be interpreted with caution in regions in which the SFCEVP is not a limiting factor such as the northwestern IP.

Hydrological droughts can be defined as the scarcity of surface and subsurface water resources that leads to negative effects in water resources management system. In this regard, our results indicate that the annual runoff seems to be a good variable to identify hydrological droughts, showing an especial good agreement with the SPEI-12.

The agreement between variables shown here also evidences that the SPEI is an adequate index to investigate drought conditions over the IP, reproducing these in a similar way than other drought-related variables over the study period. This is of high relevance to adequately develop strategies to mitigate the effects of droughts in the future.
