Jove
Visualize
Contact Us

Related Concept Videos

Precipitation Processes01:12

Precipitation Processes

The experimental conditions in a gravimetric analysis should be optimized to maximize the particle size and purity of the obtained precipitate. Ideally, the concentration of the precipitating reagent should be low with effective stirring to maintain low relative supersaturation for the growth of large crystals. In homogeneous precipitation, the precipitant is slowly generated by a chemical reaction in the solution to avoid local reagent excesses. For example, urea decomposes gradually to...
Precipitation and Co-precipitation01:17

Precipitation and Co-precipitation

Precipitation and coprecipitation methods can be used to separate a mixture of ions in a solution. In qualitative inorganic analysis, ions that form sparingly soluble precipitates with the same reagent are separated based on the differences in solubility products. For example, consider the separation of Cu(II) and Fe(II) ions by precipitation as insoluble sulfides. First, copper(II) sulfide is precipitated by the addition of acidic H2S, where the dissociation of H2S is suppressed. Adding H2S...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Comparison of the pollutant loads in dry and wet weather runoff in a southern California urban watershed.

Water science and technology : a journal of the International Association on Water Pollution Research·2002
Same author

Evaluation of land use/land cover datasets for urban watershed modeling.

Water science and technology : a journal of the International Association on Water Pollution Research·2002
See all related articles
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 11, 2026

Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling (SAHM)
12:26

Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling (SAHM)

Published on: October 11, 2016

Evaluation of an artificial neural network rainfall disaggregation model.

S J Burian1, S R Durran

  • 1Department of Civil Engineering, University of Arkansas, Fayetteville 72701, USA.

Water Science and Technology : a Journal of the International Association on Water Pollution Research
|March 13, 2002
PubMed
Summary

This study compared how well an artificial neural network model can break down hourly rainfall data into 15-minute intervals. The researchers tested if this method produces more accurate predictions of water flow in rivers than other approaches. They used a computer model of a hypothetical watershed and compared results from four different rainfall input scenarios. The neural network model's predictions had a 16.6% error in predicting peak water flow, which was better than the 40.8% error from using only hourly data and the 21.9% error from another method called geometric similarity. These results suggest the neural network model preserves important rainfall patterns better than other methods. The study shows that using artificial intelligence to refine rainfall data can improve predictions of water flow in rivers, which is important for managing water resources and predicting floods.

Keywords:
rainfall disaggregation techniqueshydrological modeling methodsartificial intelligence in meteorologyrunoff prediction accuracy

Frequently Asked Questions

Related Experiment Videos

Last Updated: Jul 11, 2026

Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling (SAHM)
12:26

Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling (SAHM)

Published on: October 11, 2016

Area of Science:

  • Hydrological modeling techniques within environmental science
  • Artificial intelligence applications in meteorology

Background:

Rainfall data at high temporal resolution is crucial for accurate hydrological simulations. Prior research has established that rainfall patterns can be disaggregated from hourly to sub-hourly intervals using artificial neural networks. However, the impact of these disaggregation errors on runoff predictions remains unclear. Existing studies have shown that rainfall-runoff models benefit from finer temporal resolution data. Yet, no prior work had resolved how specific disaggregation methods affect predicted peak discharges. This gap motivated the current evaluation of an artificial neural network model's performance in rainfall disaggregation. The study addresses the uncertainty around whether neural network-based disaggregation improves runoff predictions compared to other methods. By focusing on peak discharge errors, the research fills a specific need in hydrological modeling. The distinction between established knowledge and this study's contribution lies in the direct comparison of different disaggregation approaches. This work provides new evidence on the relative accuracy of artificial neural networks in rainfall modeling.

Purpose Of The Study:

The study aimed to evaluate how errors in rainfall disaggregation affect runoff predictions. Specifically, the researchers sought to compare the performance of an artificial neural network model against alternative methods. The primary objective was to determine whether the ANN model's rainfall patterns produce more accurate runoff hydrographs. The study focused on peak discharge errors as a key performance metric. The motivation stemmed from the need to improve hydrological modeling accuracy. The researchers hypothesized that the ANN model would outperform other disaggregation techniques. This hypothesis was tested using a hypothetical watershed model. The study's design allowed for a direct comparison of different rainfall input scenarios.

Main Methods:

The researchers used a rainfall-runoff model applied to a hypothetical watershed. They compared four different rainfall input scenarios: observed 15-minute rainfall, observed hourly rainfall, ANN-predicted 15-minute rainfall, and geometric similarity-based rainfall. The artificial neural network model had been previously trained on historical rainfall data. The model's performance was evaluated using 98 test storms. Each storm's runoff hydrograph was simulated using the four rainfall input types. Peak discharge errors were calculated for each scenario. The study focused on median under-prediction percentages as the primary outcome measure.

Main Results:

The ANN model's rainfall patterns produced a median peak discharge under-prediction of 16.6%. This was significantly better than the 40.8% under-prediction from hourly-aggregated rainfall patterns. The geometric similarity model showed a 21.9% under-prediction in peak discharge. These results suggest the ANN model outperformed both alternatives. The observed 15-minute rainfall patterns served as a benchmark for comparison. The study found consistent performance differences across all 98 test storms. The ANN model's errors were consistently lower than those of the geometric similarity model. These findings indicate the ANN model's superior ability to preserve critical rainfall characteristics.

Conclusions:

The authors concluded that the ANN model's rainfall patterns produced more accurate runoff predictions than alternative methods. The study demonstrated that the ANN model's errors were significantly lower than those from hourly-aggregated rainfall. The geometric similarity model also showed higher errors than the ANN approach. These findings suggest the ANN model preserves important rainfall characteristics better than other methods. The results support the use of artificial neural networks for rainfall disaggregation in hydrological modeling. The authors propose that this approach improves the accuracy of runoff predictions. The study's limitations include the use of a hypothetical watershed and a specific set of test storms. The findings suggest potential applications for improving flood forecasting and water resource management.

The model's rainfall patterns produced a 16.6% median under-prediction in peak discharge, better than 40.8% from hourly-aggregated rainfall.

The geometric similarity model had a 21.9% median under-prediction, higher than the ANN model's 16.6%.

A hypothetical watershed allowed direct comparison of different rainfall input scenarios without field data variability.

Median under-prediction percentages of peak discharge were calculated for each rainfall input scenario.

The study evaluated 98 test storms to compare the performance of different rainfall disaggregation methods.

The authors propose that ANN models improve runoff prediction accuracy compared to other disaggregation approaches.