India Rainfall Analytics & Predictive Modeling Dashboard

Comprehensive Hydroclimatic Evaluation (1901–2015), Regional Extremes, Seasonal Monsoon Dependency, and Next-Year Rainfall Forecasting using Linear Regression & Moving Average Models.

Historical Observations: 1901–2015 (115 Years) Meteorological Subdivisions: 36 IMD Zones District Climatological Normals: 641 Districts Data Completeness: 99.86% (70 NA imputed via Long-Term Monthly Normals) Target Forecast Year: 2016
National Long-Period Average
1,428.2 mm
Baseline annual precipitation normal across India
Wettest Subdivision (IMD)
Arunachal Pradesh
3,438.5 mm / yr (Coastal Karnataka: 3,405.9 mm)
Driest Subdivision (IMD)
West Rajasthan
292.7 mm / yr (State: Rajasthan 581.6 mm)
Peak Monsoon Dependency
Gujarat Region (95.6%)
Lowest: Tamil Nadu (35.1% - relies on Post-Monsoon)
115-Year Historical Trends
Next-Year National Forecast
1,409.8 mm
LPA Departure: -1.3% across India
Geographic Aggregation Level:
Tip: Click on any chart bar or map marker to load its 115-year historical trend.

Regional Rainfall Distribution & Extremes

Comparing long-term annual precipitation across Indian subdivisions and states

Seasonal Precipitation Composition

Breakdown across Winter (Jan-Feb), Pre-Monsoon (Mar-May), Monsoon (Jun-Sep), and Post-Monsoon (Oct-Dec)

Monsoon Dependency Ratio (%)

Proportion of annual rainfall received during the Southwest Monsoon (Jun–Sep)

Historical Trend & Next-Year Prediction Explorer (1901–2016)

Long-period regression trendline, moving-average smoothers, and next-year forecast evaluation
Long-Period Normal
-- mm
115-Year Historical Mean
Historical Range
-- - -- mm
Min & Max Recorded
115-Year Trend
--
Slope & Coefficient R²
Hydroclimatic Variability
-- %
Coefficient of Variation (CV)
Linear Regression (OLS)
Trend-based model (Y = m·X + c)
-- %
-- mm
Historical RMSE: -- mm
5-Year Moving Average
Recent short-term climatological mean
-- %
-- mm
Historical RMSE: -- mm
10-Year Moving Average
Decadal smoothed benchmark
-- %
-- mm
Historical RMSE: -- mm

Complete Regional Forecasts & Statistical Evaluation

Comprehensive data table comparing historical normals, predictive models, trends, and error metrics
Trend Filter:
Showing 36 records
Region / Zone ↑↓ Historical Normal ↑↓ Linear Regression Pred ↑↓ Departure (%) ↑↓ 5-Yr MA Pred ↑↓ 10-Yr MA Pred ↑↓ Trend ↑↓ Slope (mm/yr) ↑↓ R² Fit ↑↓ Recommended Model Action

Academic Methodology & Data Processing Notes

1. Data Cleaning & Imputation:

The 1901–2015 dataset contains 4,116 records. Out of 49,392 monthly values, 70 missing entries (0.14%) were detected and imputed using the specific subdivision's long-term monthly mean. Seasonal aggregates (Winter: Jan-Feb, Pre-Monsoon: Mar-May, Monsoon: Jun-Sep, Post-Monsoon: Oct-Dec) and annual totals were mathematically recalculated to ensure exact additive consistency.

2. Trend Analysis Classification:

Trends were determined using Ordinary Least Squares (OLS) linear regression of annual precipitation against year (1901–2015). Regions with absolute slope |m| ≥ 0.5 mm/year were classified as Increasing or Decreasing, while those with |m| < 0.5 mm/year were classified as Stable climatological regimes.

3. Predictive Modeling & Evaluation:

Next-year (2016) forecasts were generated via three methodologies: (1) OLS Linear Regression, (2) 5-Year Moving Average (recent short-term mean), and (3) 10-Year Moving Average (decadal mean). Out-of-sample one-step-ahead backtesting was performed across the historical series to compute Root Mean Squared Error (RMSE) for each model, highlighting the empirical best performer.

4. Climatological Normals & Baseline:

District normals from district_wise_rainfall_normal.csv (641 districts across 35 States/UTs) corroborate the sub-divisional long-period averages (LPA), identifying intra-state microclimatic extremes (e.g., Cherrapunji/Mawsynram in East Khasi Hills, Meghalaya vs Jaisalmer, Rajasthan).