Demand Forecasting using Historical Weather - Demo

**Overview**
At Accuweather, we know the weather. We know that it influences consumer behavior and demand as well as impacts or disrupts supply chain and logistics. There is a need to understand the signficance weather has on changes to business performance and operations. Our hope is that through prediction analysis, this same process can be applied to your own use case within your own organization with AccuWeather data.

**Scenario**
This Databricks Notebook highlights the analytical workflow of predicting the annual solar energy likely to be generated by a solar power station through AccuWeather historical information and site characteristics. The hypothesis proposed by this notebook is that a number of key weather parameters, such as temperature, wind speed, solar radiation, day length, precipitation, snowfall, and elevation will impact the daily generation of solar energy. Data is publicly available through the City of Calgary Open Data Portal with historical energy in kWh provided from October 3, 2016 through December 31, 2019 for several Solar Photovoltaic sites.

**Datasets used**
- Daily Historical Weather (AccuWeather)
- Daily Historical Solar Power ([City of Calgary](https://data.calgary.ca/Environment/Solar-Energy-Production/ytdn-2qsp))

For more details, refer to the embedded notebook.

**概述**
AccuWeather作为专业气象服务机构，深谙气象对商业活动的全方位影响：其不仅会左右消费者行为与市场需求，更会对供应链及物流环节造成冲击乃至中断。当前亟需厘清气象条件对企业经营绩效与运营变动的影响权重。我们期望通过预测分析，将基于AccuWeather气象数据的这套分析流程，落地至贵机构自身的业务场景中。

**应用场景**
本Databricks笔记本展示了一套完整的分析工作流：基于AccuWeather历史气象数据与电站场地特征，预测光伏电站的年度发电量。本笔记本提出的核心假设为：温度、风速、太阳辐射、日照时长、降水量、降雪量与海拔等关键气象参数，均会对光伏电站的单日发电量产生显著影响。相关数据可通过卡尔加里市开放数据门户公开获取，其中包含2016年10月3日至2019年12月31日期间，多个太阳能光伏（Solar Photovoltaic）站点的历史发电量数据，单位为千瓦时（kWh）。

**所用数据集**
- 历史每日气象数据（AccuWeather）
- 历史每日太阳能发电数据（[卡尔加里市](https://data.calgary.ca/Environment/Solar-Energy-Production/ytdn-2qsp)）

如需了解更多详情，请参阅内嵌笔记本。