Our idea concentrates on how to detect these floods quickly to avoid the numerous hazards. By using satellite data and maps, we created a machine learning model that detects this phenomenon and build a website application that displays the regions of the detected phenomenon, the power of near area that may be affected, and provide ancillary data to help in better understanding than by researchers and decision-makers. There a lot of features that will be available on the website also such as it will be user-friendly, the option to call for help if any viewer saw a detected region on the website, and availability of sharing the information in social media platforms to aware his community. In addition, we are developing it to be more accurate and time-saving as possible.

These features are very important to increase the efficiency of the project. In addition, it will help to avoid numerous hazards that have a great impact on us negatively caused by floods, for example, loss of crops and livestock, landslides, and deterioration of health conditions owing to waterborne diseases. As communication links and infrastructure such as power plants, roads, and bridges are damaged and disrupted, some economic activities may come to a standstill, people are forced to leave their homes and normal life is disrupted.

We developed an ML model that can predict floods and the power of the damage will occur to someplace using the data of cities of Sudan including The area of the city, and the amount of water (that causes flooding) to detect the places which will have high damage using random forest techniques in machine learning. This technique is very efficient to our system as:

• It has methods for balancing errors in data sets where classes are imbalanced.
• It handles thousands of input variables and identity most significant variables so it is considered as one of the dimensional reduction methods. Further, the model outputs the importance of variable, which can be a very handy feature.
• It has an effective method for estimating missing data and maintains accuracy when a large proportion of the data is missing.

We hope to improve this ML model and the website more to be able to include all data sets of flood all over the world to be more and more accurate. This will help a lot in avoiding a lot of accidents and save time, humans' lives, and whole communities. 

Web Application:

We made a web application to display the result of the machine learning model on it as it will be improved to display it on maps to be user friendly and very accurate in detecting places to be found easily by the viewers.

 The main purpose of the application to be a visualization for the decision makers and scientists to address and get the ancillary data of the detected regions which are predicted to be damaged by the coming flood to be studied well and used then to improve the model.

This application will be consists of some sections, the main part will be a map that its regions are colored according to some information which is helped in building the model such as water flow( high streams, medium, and low) to help in getting an overview.

There are also other sections for searching for some information concerning flood by the viewer, for example searching mainly in one county or a specific region.

It will give warnings if any predicted flood is coming to the decision-makers and specialists to help them make sure and take precautions and actions before the situation gets worse. In addition, may send some notifications for people who registered on the website to be aware of the mail (this feature will be improved as in the recommendation section).

All of this aims at the end to auto detect the flood quickly to avoid a lot of accidents form occurring.