Spatial dynamics of rainfall erosivity in the Taperoá river basin , semiarid region of Paraíba

The rainfall by rainfall erosivity is the natural element that participates more actively in soil erosion because the rain is the main factor that causes the landscape modeling and must be studied to obtain information about the intensification of the process erosion. In general, if all other factors are held constant the only variable in time and space scales are the rains making the loss of soil proportional to rainfall erosivity. Before it was necessary to evaluate the potential of rainfall erosivity in the Taperoá River basin (SBHRT), describing the spatial and temporal, which is the main objective of this work. 22 posts rainfall data were used, obtained from the DCA / UFCG site, and from them we determined the erosivity factor. Through geoprocessing tools have been created maps that showed the spatial and temporal variability of rainfall and its erosivity. All analyzes were worked through Qgis 2.14 software, and through interpolation techniques were generated maps that show the erosive potential of the basin. The results show that the SBHRT an irregular seasonal distribution of rainfall erosivity with a peak in March and April; generally the sub-basin has a high and medium erosive potential and consequently high potential for soil loss by erosion, especially in the western sector where the erosive potential is considered very high and high.


Introduction
The erosive processes are phenomen a that occur naturally in the earth's surface.The various natural elements interact with each other in order to increase or significantly soften the intensity and spread of the occurrence of soil erosion.For this, among other reasons, the individual study of each natural factor responsible for the erosion is highly necessary.
For river basins have been postulated several indicators of erosion and soil loss, among which the best known and used is the universal equation of soil loss (USLE) developed by Wishmeier and Smith in 1978, estimating the loss of soil and also allow simulation scenarios indicating usability of each of a watershed industry.
The USLE is determined from the integration of natural factors, among them the rainfall erosivity (R), soil erodibility (K), the ramp length factor and the slope (LS) and man as the use and soil management (C) and conservation practices (P).
From these natural elements, rainfall by rainfall erosivity, is the most actively participates in the soil erosion, because the rain is the main factor that causes the landscape modeling and must be studied to obtain information on the intensification of the erosion process because, in general, if all other factors are held constant the only highly variable factor in the temporal and spatial scales are the rains making the loss of proportional soil erosivity.
The definition of rainfall erosivity is simply stated by Hudson which would be the ability of rain to cause erosion (Guerra, 1998).According to Pereira (2000) with the onset of precipitation, part of the precipitate volume is somehow intercepted by vegetation or surface structures as part directly reaches the ground, moistening the aggregates and reducing its cohesion, depending on the rain duration the aggregates They disintegrate and give off later being dragged by runoff.
Rainfall is the most important factor in soil erosion and the intensity is one of its most active features.The precipitate volume and speed of the drops are directly related to the intensity and duration of rainfall (Bertoni and Lombardi Neto, 1990).Yet as important parameters for understanding the erosivity Guerra (1998) includes the timing and kinetic energy.
In this paper the study area is the subbasin of the river Taperoá (SBHRT), an important watershed in the state of Paraíba because of its strategic position in it.
The sub-basin of the river Taperoá, with an area of about 5,900 km2 is located in a region considered drier Brazil, regionally known as Cariri Paraibano, with annual rainfall between 400 and 600 mm (Lacerda, 2003;Xavier et al., 2012).Due to water scarcity and the existence of incipient soil, the basin has a low population density, and the extensive cattle the main economic activity of the region.
According to Seabra et.al. (2014), Rio Taperoá, main tributary of the Paraíba River basin is considered one of the most important and strategic position for the management of water resources in the state of Paraiba.Exactly at the confluence of the Taperoá River and the Rio Paraiba it was built a large dam forming the weir Pessoa, or simply Açude Boqueirão, due to its location in the municipality of the same name.This dam is responsible for supplying the largest metropolitan area in the Northeast, the region of Campina Grande with nearly 600 thousand in habitants.
Accordingly, the present work is general to evaluate the potential of rainfall erosivity in SBHRT, describing the spatially and temporally.
The delimitation of the basin area was performed by SRTM image taken from the site of EMBRAPA (Brazilian Agricultural Research Corporation) in GIS environment with Qgis 2.6 software lines were created following the watershed of the basin, to thus be created the perimeter the same, so it was poligonizado the basin area in shapefile format, and through the tool "Geoprocessar" >> "Cruzar" was crossed to the area obtained from the bowl with a database of the municipalities of the IBGE (Instituto Brasileiro de Geografia e estatistica ) (Figure 1) to measure the area that each municipality is the drainage basin (Table 1).

Rain characterization
For the characterization of rain basin were used monthly average data of 14 rainfall stations (Table 2), spatially distributed within the basin.For better characterization of it, all the data have an uninterrupted series of at least years and have been achieved through the official of the Department of Atmospheric Sciences at the Federal University of Campina Grande site (DCA-UFCG).The monthly data were grouped chronologically and from measures of central tendency and dispersion (standard deviation) were prepared histograms for the definition of the rainy season and more "dry" period within the basin.The rainy season was regarded as the period of at least three months which concentrate a significant percentage (> 50%) of the total precipitated during the year.
For the creation of maps of spatial distribution of rainfall were added to this initial serial data of eight other stations (Table 3) located in the same compartment geomorphologic (Borborema) above the bowl, but which nevertheless are not located within the same.
For the variable geomorphology not interfere with the space-time configuration of precipitation within the basin were despised the data from stations outside of geomorphological compartment Borborema.The summation of these data is important because of the rain be a continuous phenomenon in space and if it were carried interpolation only with data of Stations within the bowl, it could be characterized erroneously, showing a nonexistent reality.
From the tool "interpolation complement" the Qgis 2:14 geoprocessing software, the monthly rainfall data to the stations were interpolated using IDW method (Inverse power distance), it was demonstrated the spatiotemporal distribution of rainfall in bowl.With the same software were made by isohyets tool "extraction" >> "outline" for demarcating the same lines rainfall in the basin.

Erosivity rainfall
There are several methods that can be used to estimate the rainfall erosivity; The problem is to choose the most appropriate, since each event environment and are unique in spatial and temporal scales and hence erosion varies in different manners.
For Wishmeier and Smith, when all factors, except for rain, are held constant, the loss of soil by any exposed soil with unit area is directly proportional to the product of two rainfall characteristics energy kinetic (E) by its maximum intensity within 30 minutes (I 30 ) (Borges, 2009).
Several authors consider that the best method because it is noteworthy erosive characteristic of rain each 30 min event (Bertoni and lombardi Neto, 1990), however in many areas of use for is infeasible this method because of the failure and / or short collection coming from pluviographs data, and such data required to obtain the I 30.For this reason many authors attempted to correlate the factor erosivity (R) with characteristics of rainfall more easily measured and do not necessarily require intensity record, which came the following equation (1), proposed by Bertoni and Lombardi Neto (1990) used for erosivity calculations of this work."R": rain erosivity (MJ mm ha -1 ano -1 ), "p": average monthly rainfall (mm) "P" : average annual rainfall (mm).
The use of this method occurred because of that there was no satisfactory data pluviographs in SBHRT.Being a method to some simplistic point, due to the variables it uses, the method proposed by Bertoni and Lombardi Neto is subject to various criticism from some authors, because according to them it is restricted.
However, Silva and Dias (2003) to correlate the two methods showed that the difference is statistically insignificant and that, according to them, the choice of most simplistic method does not invalidate the search and therefore there is no indication that limit their use, particularly in semi-arid areas.
The average monthly data used came from 14 rain gauge stations, spatially distributed within the basin.For better characterization of it, all the data have an uninterrupted series of at least 20 years and have been achieved through the official website of the Department of Atmospheric, Federal University of Campina Grande Sciences (DCA-UFCG).
For the creation of maps of spatial distribution of erosivity were added to this initial serial data of other stations 8 located on the same geomorphological compartment (Borborema) above the bowl, but which nevertheless are not located within the same.

Rain characterization of SBHRT
Figure 2 shows the average of the monthly averages and standard deviation of rainfall, from 14 stations found within the river basin Taperoá.
It is observed that the rains are concentrated in a rainy season of three months (February, March and April) which represents more than 60% of the total precipitates during the year.In general, the SBHRT through criteria Gaussen and Bagnouls (Nimer, 1979a) all other months of the year were considered "dry" because the average temperature of the basin is around 24 ºC, with minimum values of 21 °C in from July / August and maximum of 28 °C between November / December.
In this period considered "dry" it appears that there is in the basin a very pronounced dry period of three months (September, October, November) where precipitated total do not exceed 10 mm / month.The basin has two very pronounced cycles and rain that marked the environmental characteristics of the area.The first half of the year is more rainfall, while the second half of the year precipitated values decay, giving a small increase only in the last month of the year, which characterizes the beginning of the rainy preseason next year with the month of subsequent January.
The largest standard deviations occur in the rainy season (February, March and April), showing that there is high variability of rainfall even in the rainy season.You know that historically the rain average will fluctuate naturally, so on the figures presented it can be seen that even in the rainy season may occur during the dry season, this is due to high variability of inducing rains that act not only in SBHRT but throughout Paraiba.
According to Araújo (2006) this characteristic spatial and temporal variability is due to atmospheric circulation systems that alternates form now areas / times with high rainfall sometimes areas / times with drought, as rainfall trainers mechanisms themselves throughout the area are irregular both in its spatial participation and on a time scale which causes this variability characteristic not only in the basin of river Taperoá but throughout northeastern Brazil.
According to Nimer (1979a) this variability and alternation of rain inducers in this region occur mainly by the fact that this area is the performance limit, especially the easterly waves and the ITCZ, causing heavy rains during the period of their performances and dry when not there the effect of the same.
In work carried out in the regions of Paraiba backlands, Santos (2012) found rainy and dry seasons with approximately the same duration of that found in SBHRT, however the average total annual values of precipitation per station, found in the backlands of Paraiba are relatively larger than those found in SBHRT, showing that the action of rain-inducing these areas act with greater forces than in SBHRT, confirming the assertion Nimer (1979b) From the map to the interpolated average values, prepared with the help of surrounding stations, shown in Figure 5, it is clear also that there is a greater amplitude rushed values within the basin, while the nearby city of Cabaceiras, Pocinhos and São João do Cariri annual totals are less than 400 mm and in the area corresponding to Teixeira municipalities and Cacimbas annual totals are greater than 600 mm, the western end of the basin with total close to 650 mm.
Even the greatest rainfall values found in the basin are still relatively lower than the values found, for example, by Santos (2012) in Paraiba backlands, showing that contrary to what is commonly / popularly known as the areas raining less in Northeast Brazilian.The Paraiba backlands breaks this rule at the state level, because it is easily found higher annual total to 800 mm unlike paraibano cariri where the annual totals are around 400 mm, reconfirming its status of being one of the "dry" areas Brazil.The amplitude recorded in the total annual precipitation of the Municipalities que make up the basin is due to the atmospheric movements Numerous que focus on Paraíba and cause this spatial and temporal variability.Despite being fully inserted into the semi-arid region, was Observed significant spatial variation of the total annual rainfall, where the Cabaceiras to Desterro, about 100 km straight line, will isohyet 350-650 mm respectively (Figure 5).In studies in the basin of the Paraiba River, which the basin of the Taperoá River is the main tributary, Araújo (2006) reports that several weather systems are influencers of rain on the region, with greater or lesser intensity during certain times of the year, fact which is proven in the highly irregular temporal distribution in the basin.

Erosivity rainfall
For SBHRT, the average annual rainfall varies from 304.6 to 650 mm, therefore, the test results of rainfall erosivity in the Taperoá River basin showed that the annual mean value for the basin was 3230.54MJ .mm / ha year.Through the interpolated data found a variation of 2690 MJ mm / ha year ranging from 2010.3 MJ. mm / ha year to near 4700 MJ. mm / ha.year.
The figure 4 shows the monthly average for erosivity SBHRT from the series used for the precipitation calculation.Assessing intraanual variation of erosivity indexes and correlating these values with the precipitation, we observe high coefficients of variation, the monthly average erosion rate was higher in the high rainfall period, which covers the months of March and April and lower in the months of lowest rainfall in September, October, November.
Second Frota ( 2012) is characteristic of variation of erosion rates are typical of the Brazilian northeastern semi-arid region, according to the same, the rains are characterized by short duration and high intensity, thereby generating high erosivity in this short period and increased breakdown and transport stream this calcined material.
This feature is caused by the fact that the erosivity is closely linked to the precipitated total.This behavior is expected because the erosion calculations are based on monthly and annual rainfall averages, so the greatest ability to cause erosion is the rainy season, while the lowest erosivity will be in the months of drought (Boin, 2000;Santana, 2007), which in semiarid areas can easily be null due to fact that in some years the season "dry" records null values storm or does not have significant values in order to generate erosion.Figure 5 demonstrates the values of R factor for some locations in SBHRT with direct data of rainfall stations, showing the spatial variability of erosivity.It can be seen that the difference between the places that suffer from lower and higher erosive potential of rainfall are respectively Cabaceiras and São José dos Cordeiros , and is difference increases even more after being observed values found by interpolating the data as can be observed in figure 6, where the difference between sites that have extreme potential consistent with the overall site smaller potential annual rainfall erosivity.
Figure 5-Factor R for some locations in SBHRT.
This variability of the erosive potential can also be observed in periods of the year regarded as the rainy and the dry months, showing the high variability of this factor even in the period of the same season.
Frame 1 can be seen the erosive potential of rainfall in the months that have higher rainfall during the year, there is a clear transition from erosivity index (Ei) westbound-east where in all demostrados months, the center west sector subbasin is more erosive potential in the remaining area potential remained always lower with the exception of May marks the end of the rainy season in its basin, this month the potential values are reversed and the central sectorpotential erosion, worth mentioning that even so eastern subbasin passes present the greatest the Frame 1-Months with greater erosive potential in SBHRT Frame 2 is shown the spatial distribution of rainfall erosivity in the months where the total precipitates are smaller, as can be seen due to such low values, the erosive potential is zero or very low values compared to other times of the year.
In the months where rainfall is less than the average erosive potential is 9.9 MJ. mm / ha.month ranging between 4.6 and 21.6 MJ. mm / ha.month in August and November respectively worth mentioning that beyond these amounts are very small in many areas in recent months this potential is null due to lack of rainfall.
Another fact that corroborates the high spatial variability of rainfall erosivity in semiarid areas in general, and specifically SBHRT in the municipality of Soledad found data to two rainfall seasons, but which nevertheless have between them a difference of 483.2 MJ .mm / ha year, it is noteworthy that these stations are far apart in about 8 km.
According to Figure 6 can be seen that there is downward trend of erosivity values in the west-east direction, the highest values are found in the municipality of Cacimbas with erosivity values close to 4700 MJ. mm / ha year in the same way, the municipality of Cabaceiras has the lowest erosivity values SBHRT.
Being a semi-arid region, one can see that the erosivity values (factor R) are low, showing that the rain variable in semi-arid areas compared to more humid regions, has lower potential erode soils.
Frame 2-erosive potential of rainfall for the less rainy months of the year.Other studies in the Brazilian semiarid region also showed this spatiotemporal variability of erosivity values, Aquino et al. (2006) found erosivity values of rainfall in the dry lands of the state of Piaui ranging 3316-6877 (MJ mm / ha.year) and Frota (2012) found in the state of Ceará values ranging 5137-6695 (MJ.mm / ha year), thus validating the data in SBHRT .
Table 4 shows the monthly average erosion index (EI) and the R factor for each rainfall station used, and the overall monthly mean value, standard deviation, maximum and minimum values.When comparing the values found in SBHRT of "erosivity indices" (IE) and R factor for other areas of the semi-arid northeast it is noticed that the SBHRT has low values of erosivity, incontrovertible reflections of the total annual low rainfall found in area.
To then be able to sort and compare the erosivity values with any other area, Carvalho (1994) describes reference classes for the values of R factor, as demonstrates the Table 5.  Carvalho (1994) adapted by Maciel (2014).
It is noteworthy that Carvalho (1994) Used as reference unit the metric system of erosive measures (tm.mm/ha.ano),soon to convert the metric system unit of standardized universal system (MJ.Mm / ha .year),multiply the value by 9.81 consequently to convert unit of the international system to the metric system divides the value found by 9.81.Thus, according to the reference values postulated by Carvalho (1994) in SBHRT erosion falls as weak average with predominance of the middle class, the erosivity is considered weak in the municipalities of Cabaceiras and Pocinhos and average in all other municipalities SBHRT.As can be seen in Table 6.
Table 6: erosivity class for the 14 stations inside the SBHRT Because it is universal reference values they tend to homogenize areas in a single class, not demonstrating the local intensity compared to regional erosivity values.
To this was followed by a weighting criterion similar to that found in Frota (2012) and Aquino et. al. (2006), the SBHRT was subdivided into five erosive potential classes for each class were calculated respective areas and percentages so as to identify which areas the erosion processes are more intense, this method takes into account the values found inside in the basin, so it is a uniquely made relative ranking for the same, demonstrates as Table 7. José dos Cordeiros and Livramento, caused by precipitates largest total in these two places compared to the surrounding areas.To suffer more with the erosivity problem of rains these areas must have consistent plans for environmental protection, especially of the savanna area, especially areas close to rivers, because of that vegetation is one of the natural elements that can soften the effects of erosivity (Guerra, 1998).
The vegetation acts in different ways in order to prevent the maximum action of rainfall erosivity how to avoid the direct impact of drops in the soil preventing the breakdown of particles; prevents soil compaction; increases the soil infiltration capacity and subsidize biological structures (animal and plant) that increase the permeability and porosity of the soil (Crepani, 2001).A The need for riparian forest protection in these areas are considered very high intensity, is the fact that this structure of vegetation prevents the delaminated particles of soil Durantes rainfall events from reaching the rivers, with the loss of this kills the total sediment increases to cause serious problems, particularly the dams, as well as the river itself that ends up being silted up by these sediments thus reducing its volume and capacity (Guerra, 1998).
Although the erosive intensity percentage considered "very high" is very significant in the basin, the classes "High" and "Medium" are very significant, with 31 and 38% respectively, showing that the basin naturally tends to soil loss processes water erosion reaffirming the need for management and planning in order to promote conservation practices mainly of soil management.
In the semi-arid regions, soil degradation by water erosion is a serious problem and it has been studied in recent decades by agencies and researchers who care about the issues and the socio-economic and physical relationships of the semiarid region (Santos et. al. 2007).

Conclusions
According to resultdos found, it was concluded that the sub-basin of the river Taperoá has a short rainy season of three months (February, March and April) which account for nearly 60% of the expected volume for the year.
The largest standard deviations occur in the rainy season, showing that there is high variability of rainfall even in this period.The dry season is long and remarkable in the basin, where most of the municipalities in the basin have nine months dry during the year.The rainfall data compared with other areas of Paraiba show the region as the area with the lowest rainfall in the state.
The average erosivity found was 3230.54MJ. mm / ha year.The monthly average erosion rate was higher in the high rainfall period, which covers the months of March and April and lower in the months of lowest rainfall in September, October, November.The municipality with less erosive potential, the rain is Cabaceiras and most Cacimbas.
Comparing the erosivity values of rainfall found in SBHRT with other semi-arid areas it is clear that SBHRT has low values of erosivity, incontrovertible reflections of the total annual low rainfall found in the area.The erosivity falls as weak average predominantly middle class.Spatially, the erosivity decreases in the west to east, with the municipalities of Cacimbas Texeira and the greatest potential for rainfall erosivity

Figure 1 -
Figure 1-Delimitation of SBT in the state of Paraiba and location of rainfall stations used in this study.Source: Maciel (2014).

Figure 2 -
Figure 2-Average monthly rainfall regime for the basin of Taperoá.

Figure 3 -
Figure 3-Spatial variability of the average total annual rainfall in SBHRT
this sector are lower than those recorded in the rest of SBHRT in other months of the rainy season.During the rainiest months the average erosivity values are around 716.2 MJ. mm/ ha.month, wherein the April presents the greatest potential to 1060.3 MJ. mm / ha.month and May has the potential to lower 253,8716

Table 1 -
Municipalities of the Taperoá River basin with their respective areas. .

Table 2 -
Monthly Average rainfall (mm) of the municipalities covering the river basin Taperoá.

Table 3 -
Mean rainfall data (mm) of additional stations surrounding the basin.
Source: Department of Atmospheric Sciences / UFCG

Table 4 -
rainfall stations with EI values and the factor.

Table 7 :
R factor Intervals for SBRT with their classes erosivity

Table 7 :
R factor Intervals for SBRT with their classes erosivity When analyzing Table7and Figure7, we realize that the extremes (very low and very high) occur in smaller percentage of the basin (3%).The intensities considered too low are found in the municipalities of Cabaceiras and