Predictive Hydrology in The Buturama Creek: Temporal Lag Analysis and Flow Estimation Using Hydrological Delay Models

Abstract

The key to sustainable water resource management in these parts is to forecast hydrologically. By analyzing both precipitation and streamflow through the Buturama Stream, this paper, for example, builds a predictive model for upcoming water levels based on hydrological time delays.

Through a time-delayed cross-correlation technique analysis, the paradigm has found that the precipitation change leads to fluctuations in watercourse stage with about a one to two-year period of time.

The model combines regression and historical information, and by multiregression analysis in three different parts, we have been able to produce a flow forecast (1990) for this stream. The model was tested and showed good agreement with the observed values, giving an average error of 4.15 kg/s as its Mean Absolute Error (MAE) for that statistic via ground truth, or observations, and laboratory calibration standards, 5.41 kg/s as a Root-Mean Square Error (RMSE).

While the model does have limitations--for example, during years not covered by these two terms--this flow forecasting system is fairly accurate as a general trend but unsatisfactory when extreme events are involved. Based on this, we used hypothetical precipitation data forecasts for the next decade and produced a prognosis of ten years' streamflow to find that the change on the whole is expected to be relatively slight.

Thus, the application of the present method as a forecasting technique for water resources in areas where prevalent time delays are up to five years is limited. Nevertheless, it can be successfully employed in water planning and hazard management: thus, predicting from future climate changes, we can look ahead and see what resources may become available or starved of.