SMAP Soil Moisture Analysis Pipeline

Overview

This Jupyter-based pipeline estimates soil moisture variability and anomalies over a given Area of Interest (AOI) using NASA’s SMAP (Soil Moisture Active Passive) satellite data hosted in Google Earth Engine (GEE).

It automates the process of:

1. Fetching SMAP soil moisture data from GEE.
2. Computing monthly statistics (mean, min, max) for a specified AOI.
3. Building a baseline climatology (e.g., 2016–2021).
4. Comparing recent years (2022–2024) to the baseline to compute monthly anomalies.
5. Saving all results to CSV and providing visualizations.

---

Data Sources

Dataset Used

• Name: NASA/SMAP/SPL3SMP_E.006
• Type: Enhanced Level 3 Passive Soil Moisture (9 km, Equal-Area Scalable Earth Grid)
• Temporal Coverage: 2015 – present
• Spatial Resolution: ~9 km
• Temporal Resolution: Daily global coverage
• Variables Used:
  • soil_moisture_am: Morning soil moisture (m³/m³)
  • soil_moisture_pm: Afternoon soil moisture (m³/m³)
  • Optional supporting layers (available but unused in this version):
    • tb_h_corrected_am, tb_v_corrected_am,
    • vegetation_water_content_am,
    • retrieval_qual_flag_am, etc.

Data are accessed via the Earth Engine Python API using ee.ImageCollection.

---

Pipeline Steps

1. Define AOI and Time Range

• The user specifies a polygonal Area of Interest (AOI) (e.g., administrative boundary or study region).
• Define the analysis time range (e.g., 2016–2024).

---

2. Compute Monthly SMAP Statistics

• Function: compute_monthly_smap_stats_range(aoi, start_date, end_date, pass_type='am')
• For each month in the range:
  • Filters SMAP images by date.
  • Clips to the AOI.
  • Computes the mean, min, and max soil moisture across the AOI.
• Results are stored in a DataFrame with columns:
  ['month', 'count', 'mean_soil_moisture_am', 'min_soil_moisture_am', 'max_soil_moisture_am']

---

3. Build Baseline Climatology (2016–2021)

• Aggregates monthly soil moisture statistics over the baseline years.
• Produces a climatological mean per calendar month (Jan–Dec).
• Saved as:
  outputs/smap_baseline_monthly_2016_2021_am_raw.csv

---

4. Compute Anomalies (2022–2024)

• For each target year:
  • Monthly means are computed for that year.
  • Each month’s value is compared to the baseline mean of the same month: [ \text{Anomaly} = \text{Yearly Mean} - \text{Baseline Mean} ]
  • Positive anomaly → wetter than normal.
    Negative anomaly → drier than normal.
• Results saved as: