Enabling Ecological Forecastingby integrating surface, satellite, and climate data with ecosystem models

Ramakrishna NemaniPetr VotavaAndy MichaelisForrest MeltonHirofumi HashimotoWeile WangCristina MilesiLee JohnsonLars PierceSam Hiatt

Biospheric SciencesNASA Ames Research Center

Terrestrial Observation and Prediction System

What is Ecological Forecasting?

• Ecological Forecasting (EF) predicts the effects of changes in the physical, chemical, and biological environments on ecosystem state and activity.

Short-term Monitoring and Forecasting

Sacramento river flooding, California Irrigation requirements

Based on weather forecasts, conditioned on historical ecosystem stateDays

ENSO-Rainfall over U.S

El Nino

La Nina

Based on ENSO forecastsWeeks to months

Mid-term/Seasonal Forecasts of water resources, fire risk, phenology

Long-term Projected changes

Based on GCM outputsDecades to centuries

Monitoring

Modeling

Forecasting

Multiple scales

Nemani et al., 2003, EOM White & Nemani, 2004, CJRS

A common modeling framework

Predictions are based onchanges in biogeochemicalcycles

Data – Model Integration in TOPS

TOPS-Gateway

Streamflow network Soil moisture network

FluxnetWeather network

Access to a variety of observing networks

Access to a variety of remote sensing platforms

Integration across Platforms, Sensors, Products, DAACs ..Non-trivial

Ability to integrate a variety of models

Biogeochemical CyclingCrop growth/yieldPest/Disease Global carbon cycle

Prognostic/diagnostic models

Ability to work across different time and space scales

Hours

Days

Weeks/Months

Years/Decades

Weather/Climate Forecasts at various lead timesdownscaling

Nemani et al., 2003, EOM White & Nemani, 2004, CJRS

Research & Applications of TOPS

Predictions are based onchanges in biogeochemicalcycles

Gridded Weather Surfaces for Californiausing nearly 700 weather stations daily

TMAXTMIN

VPD

SRADPRECIP

maps come with cross-validation statistics

Weather networks often operatedby different govt. agencies and/or private industry. Rarely integratedbecause they are intended fordifferent audiences. We specializein bringing them together to providespatially continuous data.

Daily satellite mapping of CA landscapes

SNOW COVER VEGETATION DENSITY

VEGETATION PHENOLOGY FIRE

California : Ecological Daily Nowcast at 1km

Biome-BGCSimulation models

Outputs include plant growth, irrigation demand, streamflowSalt water incursion, water allocation, crop coefficients

T P

RAD

Climate + Satellite Carbon and water cycles

ET

[Feb/01/2006]

0 2.5 5

GPP

GPP (gC/m2/d) ET (mm/d)

Near realtime monitoring of global NPP anomalies

Running et al., 2004, Bioscience, 54:547-560

Mapping changes in global net primary productionnear real-time depiction of the droughts in the Amazon and Horn of Africa, May 2005

0 30

Forecast Irrigation (mm)

Irrigation Forecast for week of July 19-26, 2005

Tokalon Vineyard, Oakville, CA

CIMIS Measured Weather Data through July 18, 2005

NWS Forecast Weather Data July 19-26, 2005

0 1000meters N

Irrigation Forecasts

Fully automated web delivery to growersSeasonal

•Understand the past

•Monitor/Manage the present

•Prepare for the future

Adapting TOPS for NPS needs

National Park Service

understand the past

Ecosystem changes over continental scales

understand the past

Interannual variability over Yosemite National Park

Yosemite National Park

understand the past

Watershed scale analysis of the anomalous 2004 using MODIS 250 data

Yosemite National Park

monitor the present

Snow monitoring using MODIS

Yosemite National Park

monitor the present

Monitoring stream flow

Yosemite National Park

monitor the present

Vegetation monitoring using MODIS FPAR

Yosemite National Park

monitor the present

Monitoring land surface temperature using MODIS

Yosemite National Park

prepare for the future

Impact of projected warming on Yosemite snow dynamics

Yosemite National Park

prepare for the future

Growing season dynamics under climate change

Yosemite National Park

prepare for the future

Projected trends in vegetation productivity

Yosemite National Park

Potential exists for providing ecological forecasts of various lead times

Characterizing and communicating uncertainty remains a key issue

We need:

Improved in-situ monitoring networks.

Rapid access to satellite data.

Better linkages among models.

Comprehensive framework for data management

Improved delivery systems to decision makers

Summary