**7. Results and discussion**

#### **7.1. Cannata watershed**

#### *7.1.1. Calibration test*

The observed runoff volumes from October 1996 to December 2000 at the watershed outlet were used for model calibration at monthly and event scales; annual model performance was evaluated by utilizing observations from the years 1997 to 2000. In trying to approxi‐ mate the mean and SD values of the observed runoff, the initial CNs were properly de‐ creased both in rangeland and in cropland areas (Table 3). Table 5 shows the values of the chosen difference measures obtained for runoff at annual, monthly and event scales before and after calibration.


**Coefficient or measure Equation Range of**

*n*

=

2 1

r

Root Mean Square Error ( )

= observed and predicted values at the time step i.

**Table 4.** Coefficients and difference measures for model evaluation and their range of variability.

=

( )( )

*O OPP*

é ù ê ú

( ) ( )

( )

*O P*

( )

*O P*

*O O*

2

<sup>=</sup> <sup>0</sup> to <sup>∞</sup>

*O O*

*O O PP*

ë û

å -

å -

1

å -

å -

1 1

*O P*

*O* = =

*n n i i i i n <sup>i</sup> <sup>i</sup>*

å å -

1

=

å

=

*n i i <sup>i</sup> n <sup>i</sup> <sup>i</sup>*

1

=

*n i i <sup>i</sup> P O n*

= å -

1 1

= =

*i i <sup>i</sup> n n i i i i*

å å - -

1

=

*n i i <sup>i</sup> n <sup>i</sup> <sup>i</sup>*

1

= -

1

*E*

1

= -

<sup>1</sup> RMSE

CRM

The observed runoff volumes from October 1996 to December 2000 at the watershed outlet were used for model calibration at monthly and event scales; annual model performance was evaluated by utilizing observations from the years 1997 to 2000. In trying to approxi‐ mate the mean and SD values of the observed runoff, the initial CNs were properly de‐ creased both in rangeland and in cropland areas (Table 3). Table 5 shows the values of the chosen difference measures obtained for runoff at annual, monthly and event scales before

=

*E*

1

=

å - -

2 2

2

2

Coefficient of determination

Coefficient of efficiency (Nash and

Modified coefficient of efficiency

Coefficient of residual mass (Loague

n = number of observations.

Ō = mean of observed values.

Sutcliffe, 1970)

16 Research on Soil Erosion Soil Erosion

(Willmott, 1982)

and Green, 1991)

Oi , Pi

**7.1. Cannata watershed**

*7.1.1. Calibration test*

and after calibration.

**7. Results and discussion**

**variability**

0 to 1

−∞ to 1

−∞ to 1

−∞ to ∞

2

**Table 5.** Values of the coefficients, summary and difference measures applied to runoff volumes at different time scales for calibration and validation tests at the Cannata watershed.

The simulated total runoff volume for the period of October 1996 to December 2000 (405.72 mm) was only slightly higher than the observed value (393.23 mm), showing a runoff pre‐ diction capability for long periods, which was also detected by other Authors (Yuan et al., 2001). The improvement in the annual runoff volume predictions after the calibration is due to the reduction of the cumulated volume overprediction relative to events with smaller runoff (Figure 5). In some cases, at the beginning of the wet season, runoff was generated by AnnAGNPS but not observed (Figure 6). This was probably due to the peculiarity of the hy‐ drological processes governing runoff formation in Mediterranean regions, depending not only on catchment characteristics but also on antecedent hydrological conditions and char‐ acteristics of the rainfall events, with low runoff coefficients as a result of short-duration, high-intensity convective storms over dry soils (Latron et al., 2003).

of E1 and the low RMSE and CRM (Table 5). A similar value of E was found in the model

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model

http://dx.doi.org/10.5772/50427

19

The apparent best results achieved for monthly and event-scale runoff volume predictions with respect to annual values may depend on the fact that the simulation period only repre‐ sents a few years of data (four years and three years for the calibration and validation peri‐ ods, respectively), while monthly and event-scale simulations provide more data for the statistics. Moreover, in Table 5, results of simulations related to the period of October to De‐

**Figure 7.** Comparison between observed and simulated runoff at event scale for (left) calibration and (right) valida‐

As expected, the coefficient E1 is less sensitive to peaks (Krause et al., 2005) and was general‐

Adjustments of minimum and maximum interception evaporation (the portion of precipita‐ tion that neither runs off nor infiltrates) within the lower and upper default bounds as‐ sumed by AnnAGNPS for daily pluviometric and meteorological data did not improve the

Peak flow predictions were closer to the observed values when the type Ia synthetic 24-h rainfall distribution (less intense than type I) was used. The overall model performance was

High values of the coefficient of determination and model efficiency (E and E1) were found for the suspended sediment yield events observed from October 1996 to December 2000 (Figure 8) when the AnnAGNPS erosive submodel was calibrated (Table 7). By decreasing the surface long-term random roughness coefficient as well as the sheet and concentrated flow Manning's roughness coefficients for both rangeland and cropland areas, the tendency to underprediction was substantially reduced. The model response was remarkably more

ly lower than E, but nevertheless satisfactory after the calibration process.

satisfactory for less intense events, as shown by the E1 coefficient (Table 6).

cember 1996, which was very well simulated by the model, are not reported.

calibration test reported by Baginska et al. (2003).

tion tests at the Cannata watershed.

model prediction capability.

**Figure 5.** Comparison between observed and simulated (using default and calibrated parameters) yearly runoff vol‐ ume for the years 1997 to 2003 at the Cannata watershed.

**Figure 6.** Comparison between observed and simulated (using default and calibrated parameters) monthly runoff vol‐ ume for the whole period at the Cannata watershed.

The goodness of fit between observed and simulated runoff volumes (Figure 7) was also confirmed at the event scale by the summary measures as well as by the satisfactory values of E1 and the low RMSE and CRM (Table 5). A similar value of E was found in the model calibration test reported by Baginska et al. (2003).

The apparent best results achieved for monthly and event-scale runoff volume predictions with respect to annual values may depend on the fact that the simulation period only repre‐ sents a few years of data (four years and three years for the calibration and validation peri‐ ods, respectively), while monthly and event-scale simulations provide more data for the statistics. Moreover, in Table 5, results of simulations related to the period of October to De‐ cember 1996, which was very well simulated by the model, are not reported.

**Figure 5.** Comparison between observed and simulated (using default and calibrated parameters) yearly runoff vol‐

**Figure 6.** Comparison between observed and simulated (using default and calibrated parameters) monthly runoff vol‐

The goodness of fit between observed and simulated runoff volumes (Figure 7) was also confirmed at the event scale by the summary measures as well as by the satisfactory values

ume for the years 1997 to 2003 at the Cannata watershed.

18 Research on Soil Erosion Soil Erosion

ume for the whole period at the Cannata watershed.

**Figure 7.** Comparison between observed and simulated runoff at event scale for (left) calibration and (right) valida‐ tion tests at the Cannata watershed.

As expected, the coefficient E1 is less sensitive to peaks (Krause et al., 2005) and was general‐ ly lower than E, but nevertheless satisfactory after the calibration process.

Adjustments of minimum and maximum interception evaporation (the portion of precipita‐ tion that neither runs off nor infiltrates) within the lower and upper default bounds as‐ sumed by AnnAGNPS for daily pluviometric and meteorological data did not improve the model prediction capability.

Peak flow predictions were closer to the observed values when the type Ia synthetic 24-h rainfall distribution (less intense than type I) was used. The overall model performance was satisfactory for less intense events, as shown by the E1 coefficient (Table 6).

High values of the coefficient of determination and model efficiency (E and E1) were found for the suspended sediment yield events observed from October 1996 to December 2000 (Figure 8) when the AnnAGNPS erosive submodel was calibrated (Table 7). By decreasing the surface long-term random roughness coefficient as well as the sheet and concentrated flow Manning's roughness coefficients for both rangeland and cropland areas, the tendency to underprediction was substantially reduced. The model response was remarkably more sensitive to the random roughness (more than 95% of the model efficiency improvement) than the Manning's coefficients adjustments (Table 3).


**Figure 8.** Comparison between observed and simulated sediment yield at event scale for (left) calibration and (right)

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model

http://dx.doi.org/10.5772/50427

21

The performance of the calibrated model was evaluated for the period of January 2001 to

AnnAGNPS runoff volume predictions confirmed the satisfactory model performance both at the event and annual scales and the good performance at the monthly aggregated values (Ta‐ ble 5). However, an underprediction was highlighted by the difference in summary measures and the values of RMSE and CRM. This tendency was mainly due to underestimation of the more significant events (Figure 7), as also found in the tests performed by Yuan et al. (2001).

The poor performance of the model in predicting extreme peak flows was confirmed in the val‐ idation period. The overall model prediction capability was unsatisfactory (Table 6), as shown by the poor value of the coefficient of efficiency (E = 0.05). A high overprediction (over 105%)

were also found for the suspended sediment yield events observed in the period of 2001 to 2003 (Table 7 and Figure 8). The satisfactory value achieved for the Nash and Sutcliffe coefficient (E = 0.87) was mainly due to the successful performance of the model for large rainfall events, in

Runoff depths were in general underpredicted (see the positive value of the CRM coefficient in Table 8). The accuracy achieved for the prediction of the largest event (13-14/09/1998) gave a coefficient of determination exceeding 0.90 (Figure 9) and a model efficiency (E) of 0.89 for runoff depth (Table 8). The mean and standard deviation of simulated runoff vol‐

> 0.90)

for the most significant event, which occurred on 12 December 2003, is also noted.

particular for the highest sediment yield, which occurred on 12 December 2003.

A satisfactory model efficiency (E1 = 0.55) and a very high coefficient of determination (r2

December 2003 in terms of runoff, peak flow and sediment yield.

validation tests at the Cannata watershed.

*7.1.2. Validation test*

**7.2. Ganspoel watershed**

**Table 6.** Values of the coefficients, summary and difference measures applied to peak flow at event scale for calibration and validation tests at the Cannata watershed.

Peak flow and sediment yield predictions were only slightly sensitive to the calibration of the hydrological submodel; the model efficiency in sediment yield prediction did not in‐ crease by adjusting either the Manning's roughness coefficient for channels or the ratio of rill to inter-rill erosion for bare soil.


**Table 7.** Values of the coefficients, summary and difference measures applied to sediment yield at event scale for calibration and validation tests at the Cannata watershed.

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model http://dx.doi.org/10.5772/50427 21

**Figure 8.** Comparison between observed and simulated sediment yield at event scale for (left) calibration and (right) validation tests at the Cannata watershed.

#### *7.1.2. Validation test*

sensitive to the random roughness (more than 95% of the model efficiency improvement)

**(m3 s-1) r2 <sup>E</sup> E1**

*Calibration test (Oct. 1996 to Dec. 2000)*

Predicted [a] 0.03 0.33 0.57 -4.04 0.05 0.26 -1.12 Predicted b] 0.01 0.14 0.56 0.34 0.52 0.09 0.14 *Validation test (Jan. 2001 to Dec. 2003)*

Predicted[b] 0.02 0.23 0.66 0.05 0.51 0.14 0.11

Peak flow and sediment yield predictions were only slightly sensitive to the calibration of the hydrological submodel; the model efficiency in sediment yield prediction did not in‐ crease by adjusting either the Manning's roughness coefficient for channels or the ratio of rill

**(Mg) r2 <sup>E</sup> E1**

*Calibration test (Oct. 1996 to Dec. 2000)* Observed 23.31 28.30 -- -- -- -- -- Predicted [a] 11.00 16.46 0.84 0.51 0.49 18.52 0.53 Predicted [b] 17.16 25.74 0.84 0.79 0.71 12.27 0.26 *Validation test (Jan. 2001 to Dec. 2003)* Observed 26.17 69.13 -- -- -- -- -- Predicted [b] 32.14 81.62 0.92 0.87 0.55 24.34 -0.23

**Table 7.** Values of the coefficients, summary and difference measures applied to sediment yield at event scale for

**Table 6.** Values of the coefficients, summary and difference measures applied to peak flow at event scale for

**RMSE (m3 s-1)**

> **RMSE (Mg)**

**CRM**

**CRM**

than the Manning's coefficients adjustments (Table 3).

Observed 0.02 0.11

Observed 0.02 0.14

[a] Default simulation [b] Calibrated model.

calibration and validation tests at the Cannata watershed.

to inter-rill erosion for bare soil.

**Values**

[a] Default simulation [b] Calibrated model.

calibration and validation tests at the Cannata watershed.

**Mean (Mg)**

**Std. Dev.**

**Mean (m3 s-1)** **Std. Dev.**

**Values**

20 Research on Soil Erosion Soil Erosion

The performance of the calibrated model was evaluated for the period of January 2001 to December 2003 in terms of runoff, peak flow and sediment yield.

AnnAGNPS runoff volume predictions confirmed the satisfactory model performance both at the event and annual scales and the good performance at the monthly aggregated values (Ta‐ ble 5). However, an underprediction was highlighted by the difference in summary measures and the values of RMSE and CRM. This tendency was mainly due to underestimation of the more significant events (Figure 7), as also found in the tests performed by Yuan et al. (2001).

The poor performance of the model in predicting extreme peak flows was confirmed in the val‐ idation period. The overall model prediction capability was unsatisfactory (Table 6), as shown by the poor value of the coefficient of efficiency (E = 0.05). A high overprediction (over 105%) for the most significant event, which occurred on 12 December 2003, is also noted.

A satisfactory model efficiency (E1 = 0.55) and a very high coefficient of determination (r2 > 0.90) were also found for the suspended sediment yield events observed in the period of 2001 to 2003 (Table 7 and Figure 8). The satisfactory value achieved for the Nash and Sutcliffe coefficient (E = 0.87) was mainly due to the successful performance of the model for large rainfall events, in particular for the highest sediment yield, which occurred on 12 December 2003.

#### **7.2. Ganspoel watershed**

Runoff depths were in general underpredicted (see the positive value of the CRM coefficient in Table 8). The accuracy achieved for the prediction of the largest event (13-14/09/1998) gave a coefficient of determination exceeding 0.90 (Figure 9) and a model efficiency (E) of 0.89 for runoff depth (Table 8). The mean and standard deviation of simulated runoff vol‐ ume depths were close to the corresponding observed values with differences lower than 12% and 16%. When the events for which zero runoff was simulated events were excluded from the analysis, the values for r2 and E become 0.98 and 0.97 respectively. Similarly high values for the coefficient of determination were found for runoff simulations by AnnAGNPS at the event scale by Yuan et al. (2001), Shrestha et al. (2006) and Shamshad et al. (2008) and for the coefficient of determination and model efficiency by Sarangi et al. (2007). However, in these studies AnnAGNPS was calibrated before a validation was carried out.

and predicted values were of the same order of magnitude (Table 8), the low values ach‐ ieved by the coefficients of efficiency (E and E1 lower than 0.35) and conversely the high RMSE (163% of observed mean, Table 8) utilized for model evaluation confirmed the unsat‐ isfactory prediction capability of the model for peak flow, also found elsewhere in different model tests (Shrestha et al., 2006). The model uses the extended TR-55 methods through syn‐ thetic 24-h rainfall distributions to calculate the peak flow (Cronshey and Theurer, 1998). Apparently, the latter method results is not suitable for the study area, leading to a severe underestimation of rainfall intensities and hence peak flows, a fact also noted by Shrestha et al. (2006). A prediction method that takes into account the actual patterns of rainfall intensi‐

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model

**Runoff**

**RMSE (mm)**

**RMSE (m3 s-1)**

**RMSE (Mg)**

**CRM**

http://dx.doi.org/10.5772/50427

23

**CRM**

**CRM**

**(mm) r2 <sup>E</sup> E1**

Observed 1.04 2.26 - - - - - Predicted 0.87 2.53 0.92 0.89 0.59 0.73 0.16 **Peak flow**

**(m3 s-1) r2 <sup>E</sup> E1**

**(Mg) r2 <sup>E</sup> E1**

Predicted sediment yields were strongly underestimated with respect to the observed values (up to one order of magnitude in three cases); the correlation between observed and predict‐ ed values was relatively low (Table 9; Figure 9). Coefficients of efficiency (E and E1) were close to zero and the coefficient of determination did not exceed 0.60 (Table 8). Those results were in accordance of what reported by Yuan et al. (2001), Shrestha et al. (2006), Polyakov et

The model tendency to strongly underpredict peak flow is probably one of the main rea‐ sons for the underestimation of erosive events and, consequently, of sediment yield (also

Observed 8.54 17.65 - - - - - Predicted 1.84 4.31 0.57 0.16 0.29 15.71 0.78

Observed 0.16 0.30 - - - - - Predicted 0.12 0.39 0.53 0.35 0.19 0.26 0.27 **Sediment yield**

ty would be expected to provide better accuracy in peak flow estimations.

**Std. Dev.**

**Std. Dev.**

**Std. Dev.**

**Table 8.** Statistics concerning the AnnAGNPS simulations of 17 events at the Ganspoel watershed.

al. (2007) and Shamshad et al. (2008) in sediment yield modeling by AnnAGNPS.

**Values**

**Mean (mm)**

**Mean (m3 s-1)**

**Mean (Mg)**

From such outcomes it can be remarked that the AnnAGNPS model provided a generally good capability to simulate the greatest runoff event in the Ganspoel watershed, as shown by the high coefficients of efficiency (E and E1) and determination (r2 ) achieved without any a priori calibration. The latter is an important observation as it shows that, at least for signif‐ icant events, adequate runoff modeling is possible without calibration provided that suffi‐ ciently detailed input data are available. The latter should not only contain land use, but also surface characteristics and soil roughness as these are important controls on runoff pro‐ duction. This result contrasts somewhat with that of many other studies, where the need for appropriate calibration is stressed (e.g. Refsgaard, 1997; Beven, 2006). A possible reason for this is that in many cases the available input data are less detailed than those available for the Ganspoel watersheds in terms of soil surface characteristics and coverage. The latter are important controls on runoff generation: if such data are not available, model predictions cannot be expected to be accurate without prior calibration.

The majority of the observations available in the hydrological database was of low magni‐ tude (14 out of 17 with runoff depths lower than 1 mm); for them the model simulation ac‐ curacy was basically less accurate, achieving a mean deviation between simulations and observations of about 50%. Moreover, seven events (five of them concentrated at the end of relatively dry periods and generated by storms with a depth up to 13 mm) resulted in zero runoff simulation, even tuning the values of the initial CNs or saturated hydraulic conduc‐ tivity (which represent the most important input parameters to which the runoff is sensitive (Yuan et al., 2001; Baginska et al., 2003) and setting up pre-run before the first event simulat‐ ed (which is important for initial soil moisture). The AnnAGNPS model, calculating daily and sub-daily water budgets using NRCS TR-55 method coming from the SWRRB and EPIC models (Williams et al, 1984; USDA-NRCS, 1986), presumably would have adjusted the CNs to antecedent moisture condition AMC-I based on the NRCS criteria, minimising the effect of varying the CNs (Sarangi et al., 2007). The climatic characteristics of the studied water‐ shed caused the model to produce unrealistic CN values during its initialization and, as a result, too low or no predicted runoff, as also found in various experimental applications in different climatic conditions (Polyakov et al., 2007; Sarangi et al., 2007).

Even in the Ganspoel watershed adjustments of minimum and maximum interception, as operated for model's implementation at the Cannata watershed, did not further improve the coefficients E, E1 and r2 calculated for runoff volume prediction.

The AnnAGNPS model provided the highest accuracy in peak flow predictions when the type "II" synthetic 24-h rainfall distribution (typical of continental climate, with cold winter and warm summer) was set in simulation tests (Figure 9). Even though statistics of observed and predicted values were of the same order of magnitude (Table 8), the low values ach‐ ieved by the coefficients of efficiency (E and E1 lower than 0.35) and conversely the high RMSE (163% of observed mean, Table 8) utilized for model evaluation confirmed the unsat‐ isfactory prediction capability of the model for peak flow, also found elsewhere in different model tests (Shrestha et al., 2006). The model uses the extended TR-55 methods through syn‐ thetic 24-h rainfall distributions to calculate the peak flow (Cronshey and Theurer, 1998). Apparently, the latter method results is not suitable for the study area, leading to a severe underestimation of rainfall intensities and hence peak flows, a fact also noted by Shrestha et al. (2006). A prediction method that takes into account the actual patterns of rainfall intensi‐ ty would be expected to provide better accuracy in peak flow estimations.

ume depths were close to the corresponding observed values with differences lower than 12% and 16%. When the events for which zero runoff was simulated events were excluded

values for the coefficient of determination were found for runoff simulations by AnnAGNPS at the event scale by Yuan et al. (2001), Shrestha et al. (2006) and Shamshad et al. (2008) and for the coefficient of determination and model efficiency by Sarangi et al. (2007). However,

From such outcomes it can be remarked that the AnnAGNPS model provided a generally good capability to simulate the greatest runoff event in the Ganspoel watershed, as shown

a priori calibration. The latter is an important observation as it shows that, at least for signif‐ icant events, adequate runoff modeling is possible without calibration provided that suffi‐ ciently detailed input data are available. The latter should not only contain land use, but also surface characteristics and soil roughness as these are important controls on runoff pro‐ duction. This result contrasts somewhat with that of many other studies, where the need for appropriate calibration is stressed (e.g. Refsgaard, 1997; Beven, 2006). A possible reason for this is that in many cases the available input data are less detailed than those available for the Ganspoel watersheds in terms of soil surface characteristics and coverage. The latter are important controls on runoff generation: if such data are not available, model predictions

The majority of the observations available in the hydrological database was of low magni‐ tude (14 out of 17 with runoff depths lower than 1 mm); for them the model simulation ac‐ curacy was basically less accurate, achieving a mean deviation between simulations and observations of about 50%. Moreover, seven events (five of them concentrated at the end of relatively dry periods and generated by storms with a depth up to 13 mm) resulted in zero runoff simulation, even tuning the values of the initial CNs or saturated hydraulic conduc‐ tivity (which represent the most important input parameters to which the runoff is sensitive (Yuan et al., 2001; Baginska et al., 2003) and setting up pre-run before the first event simulat‐ ed (which is important for initial soil moisture). The AnnAGNPS model, calculating daily and sub-daily water budgets using NRCS TR-55 method coming from the SWRRB and EPIC models (Williams et al, 1984; USDA-NRCS, 1986), presumably would have adjusted the CNs to antecedent moisture condition AMC-I based on the NRCS criteria, minimising the effect of varying the CNs (Sarangi et al., 2007). The climatic characteristics of the studied water‐ shed caused the model to produce unrealistic CN values during its initialization and, as a result, too low or no predicted runoff, as also found in various experimental applications in

Even in the Ganspoel watershed adjustments of minimum and maximum interception, as operated for model's implementation at the Cannata watershed, did not further improve the

The AnnAGNPS model provided the highest accuracy in peak flow predictions when the type "II" synthetic 24-h rainfall distribution (typical of continental climate, with cold winter and warm summer) was set in simulation tests (Figure 9). Even though statistics of observed

calculated for runoff volume prediction.

in these studies AnnAGNPS was calibrated before a validation was carried out.

by the high coefficients of efficiency (E and E1) and determination (r2

cannot be expected to be accurate without prior calibration.

different climatic conditions (Polyakov et al., 2007; Sarangi et al., 2007).

coefficients E, E1 and r2

and E become 0.98 and 0.97 respectively. Similarly high

) achieved without any

from the analysis, the values for r2

22 Research on Soil Erosion Soil Erosion


**Table 8.** Statistics concerning the AnnAGNPS simulations of 17 events at the Ganspoel watershed.

Predicted sediment yields were strongly underestimated with respect to the observed values (up to one order of magnitude in three cases); the correlation between observed and predict‐ ed values was relatively low (Table 9; Figure 9). Coefficients of efficiency (E and E1) were close to zero and the coefficient of determination did not exceed 0.60 (Table 8). Those results were in accordance of what reported by Yuan et al. (2001), Shrestha et al. (2006), Polyakov et al. (2007) and Shamshad et al. (2008) in sediment yield modeling by AnnAGNPS.

The model tendency to strongly underpredict peak flow is probably one of the main rea‐ sons for the underestimation of erosive events and, consequently, of sediment yield (also shown by the separate comparison of deposition and erosion values for observed and si‐ mulated events, Van Oost et al., 2005), but is not the only one. Also in the case of a good estimation of the runoff volume and an overestimation of the peak flow (13-14/09/1998), the sediment yield was underestimated. Runoff alone is not adequate for erosion and sediment delivery predictions, but in the AnnAGNPS erosion sub-model it is used to esti‐ mate the delivery of the particle sizes of eroded sediment (simulated through the RUSLE model) based on runoff and peak flow.

**Date**

watershed.

dicted sediment yields:

**Rainfall**

**Runoff volume**

**Peak flow Sediment yield**

http://dx.doi.org/10.5772/50427

25

**Obs. Sim. Obs. Sim. Observed Simulated**

(mm) (mm) (m3 s-1) (Mg) (kg ha-1) (Mg) (kg ha-1)

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model

19/05/1997 8.0 0.22 0 0.103 0 8.20 70.09 0 0 21/05/1997 6.5 0.13 0 0.056 0 2.73 23.32 0 0 11/07/1997 13.0 1.97 0 0.862 0 40.91 349.68 0 0 14/07/1997 5.5 0.37 0 0.181 0 4.40 37.63 0 0 17-18/07/1997 21.5 0.35 0.04 0.050 0.003 3.60 30.78 2.18 18.63 25/12/1997 6.5 0.09 0.20 0.043 0.032 0.25 2.11 0.01 0.09 05/01/1998 8.0 0.23 0.21 0.051 0.034 0.53 4.53 0.01 0.09 28/04/1998 11.0 0.14 0 0.037 0 0.21 1.76 0 0 26/08/1998 5.5 0.39 0 0.064 0 1.89 16.18 0 0 08-09/09/1998 24.5 0.45 0.60 0.067 0.023 1.30 11.09 4.17 35.64 13-14/09/1998 57.5 8.86 10.55 1.017 1.629 66.13 565.19 17.66 150.94 31/10-01/11/1998 25.0 1.67 0.01 0.064 0.062 6.89 58.92 0 0 14/11/1998 15.5 0.71 0.68 0.032 0.038 0.72 6.13 2.61 22.31 29/11/1998 18.5 0.56 1.62 0.025 0.129 1.40 12.01 3.51 30.00 16-17/01/1999 14.5 0.94 0.73 0.033 0.047 2.55 21.80 1.09 9.32 28/01/1999 8.0 0.71 0.11 0.046 0.029 2.99 25.57 0 0 07/02/1999 6.5 0.30 0 0.029 0 0.55 4.69 0 0

**Table 9.** Main characteristics of the observed events and simulations by the AnnAGNPS model at the Ganspoel

Moreover, the following factors can explain the low correlation between observed and pre‐

**•** AnnAGNPS uses the RUSLE method as the erosion sub-model. RUSLE has been devel‐ oped to deliver estimates of long-term average erosion rates rather than event-based sim‐ ulations. For this reason, comparison of individual events may not agree as well as longterm annual values (Shrestha et al., 2006), even in the case of adequate prediction for the

**•** we deliberately opted to evaluated the AnnAGNPS model without prior validation in or‐

**•** the Ganspoel watershed contains more than 80 fields (roads, buildings, forest, grassed chan‐ nels and several crops with differing planting and harvesting schedules), showing difficul‐ ties for modeling of interactions between physical processes (water evapotranspiration,

der to assess its performance in cases where no data for validation are available;

most intense runoff events, as achieved in our model tests;

**Figure 9.** Comparison of 17 observed and simulated (by AnnAGNPS) events in the Ganspoel watershed, for runoff (upper left), peak flow (upper right) and sediment yield (bottom) (values are in logarithmic scale).

However, another factor that may also play a role in poor model simulations of erosion was the limited availability of input parameters. The AnnAGNPS model requires up to 100 unique parameters for runoff volume assessment and up to an additional 80 unique parame‐ ters for sediment yield prediction. As values for these parameters were not all available in the Ganspoel dataset data from the literature had to be used in some cases.


shown by the separate comparison of deposition and erosion values for observed and si‐ mulated events, Van Oost et al., 2005), but is not the only one. Also in the case of a good estimation of the runoff volume and an overestimation of the peak flow (13-14/09/1998), the sediment yield was underestimated. Runoff alone is not adequate for erosion and sediment delivery predictions, but in the AnnAGNPS erosion sub-model it is used to esti‐ mate the delivery of the particle sizes of eroded sediment (simulated through the RUSLE

**Figure 9.** Comparison of 17 observed and simulated (by AnnAGNPS) events in the Ganspoel watershed, for runoff

However, another factor that may also play a role in poor model simulations of erosion was the limited availability of input parameters. The AnnAGNPS model requires up to 100 unique parameters for runoff volume assessment and up to an additional 80 unique parame‐ ters for sediment yield prediction. As values for these parameters were not all available in

(upper left), peak flow (upper right) and sediment yield (bottom) (values are in logarithmic scale).

the Ganspoel dataset data from the literature had to be used in some cases.

model) based on runoff and peak flow.

24 Research on Soil Erosion Soil Erosion

**Table 9.** Main characteristics of the observed events and simulations by the AnnAGNPS model at the Ganspoel watershed.

Moreover, the following factors can explain the low correlation between observed and pre‐ dicted sediment yields:


interception, infiltration and runoff as well as soil detachment and transport) and water and sediment routing associated with its complexity (Nearing et al., 2005; Licciardello et al., 2009). Probably, the scale of soil property measurements within the available geomorpho‐ logical database does not correspond to the discretisation scale of the Ganspoel watershed (characterized by land use heterogeneity and crop schedule complexity, as mentioned above) performed by the GIS interface of the data-intensive AnnAGNPS model.

availability of geomorphologic and hydrologic information within the experimental data‐ base as well as the easiness of finding/measuring the majority of input parameters (e.g. me‐ teorological data, soil physical properties). Nevertheless, in some cases processing of simulated hydrologic variables resulted in a time consuming task, especially for surface run‐

Prediction of Surface Runoff and Soil Erosion at Watershed Scale: Analysis of the AnnAGNPS Model

The model performance could be further improved by optimising algorithms for water bal‐ ance of soil (in order to improve the simulation of more realistic moisture conditions) or by utilising as input the observed rainfall patterns (at hourly or sub-hourly scales) instead of the synthetic hyetographs utilised at present by AnnAGNPS. Sensitivity analyses, which would allow a more precise estimation of the input parameters to which model response is

Such improvements, together further research activities aiming at model verification in dif‐ ferent environmental conditions, could enhance the model consolidation and stimulate its wider diffusion in professional activities for controlling surface runoff and soil erosion as

1 Mediterranean University of Reggio Calabria, Department of Agro-forest and Enviromen‐

2 University of Catania, Department of Agrofood and Environemental System Management,

[1] Aguilar, C., & Polo, M. J. (2005). Análisis de sensibilidad de AnnAGNPS en la diná‐ mica de herbicidas en cuencas de olivar. *In: FJ Samper Calvete y A Paz González, editors.*

[2] AnnAGNPS version 2 user documentation (2001). Available: http://

**Notes:** The contributions of the authors to this work can be considered equivalent.

*Estudios de la Zona No Saturada del Suelo*, VII, La Coruna, Spain.

www.ars.usda.gov Accessed 2007 Jan 22.

, Pietro Denisi1

, Feliciana Licciardello2

http://dx.doi.org/10.5772/50427

27

and

more sensitive, would be advisable for a better model implementation.

well as planning mitigation countermeasures.

Demetrio Antonio Zema1\*, Giuseppe Bombino1

\*Address all correspondence to: dzema@unirc.it

off analysis at event scale.

**Author details**

Italy

**References**

Santo Marcello Zimbone1

tal Science and Technology, Italy
