SEARCH · Engineering Papers

Results for “climate data”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,639 records · Page 91

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Data Imbalance, Uncertainty Quantification, and Transfer Learning in Data‐Driven Parameterizations: Lessons From the Emulation of Gravity Wave Momentum Transport in WACCM

Abstract Neural networks (NNs) are increasingly used for data‐driven subgrid‐scale parameterizations in weather and climate models. While NNs are powerful tools for learning complex non‐linear relationships from data, there are several challenges in using them for parameterizations. Three of these challenges are (a) data imbalance related to learning rare, often large‐amplitude, samples; (b) uncertainty quantification (UQ) of the predictions to provide an accuracy indicator; and (c) generalization to other climates, for example, those with different radiative forcings. Here, we examine the performance of methods for addressing these challenges using NN‐based emulators of the Whole Atmosphere Community Climate Model (WACCM) physics‐based gravity wave (GW) parameterizations as a test case. WACCM has complex, state‐of‐the‐art parameterizations for orography‐, convection‐, and front‐driven GWs. Convection‐ and orography‐driven GWs have significant data imbalance due to the absence of convection or orography in most grid points. We address data imbalance using resampling and/or weighted loss functions, enabling the successful emulation of parameterizations for all three sources. We demonstrate that three UQ methods (Bayesian NNs, variational auto‐encoders, and dropouts) provide ensemble spreads that correspond to accuracy during testing, offering criteria for identifying when an NN gives inaccurate predictions. Finally, we show that the accuracy of these NNs decreases for a warmer climate (4 × CO 2 ). However, their performance is significantly improved by applying transfer learning, for example, re‐training only one layer using ∼1% new data from the warmer climate. The findings of this study offer insights for developing reliable and generalizable data‐driven parameterizations for various processes, including (but not limited to) GWs.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2020

Alaskan carbon-climate feedbacks will be weaker than inferred from short-term manipulations: Alaskan Benchmark Data and Model runs

This submission aimed to assess differences in short-term step warming manipulations and long-term chronic response to climate change in Alaskan ecosystems. Briefly, climate warming is occurring fastest at high latitudes. Based on short-term field experiments, this warming is projected to stimulate soil organic matter decomposition, and promote a positive feedback to climate change. We show here that the tightly coupled, nonlinear nature of high-latitude ecosystems implies that short-term (< 10 year) warming experiments produce emergent ecosystem carbon stock temperature sensitivities inconsistent with emergent multi-decadal responses. We first demonstrate that a well-tested mechanistic ecosystem model accurately represents observed carbon cycle and active layer depth responses to short-term summer warming in four diverse Alaskan sites. We then show that short-term warming manipulations do not capture the non-linear, long-term dynamics of vegetation, and thereby soil organic matter, that occur in response to thermal, hydrological, and nutrient transformations belowground. Our results demonstrate significant spatial heterogeneity in multi-decadal Arctic carbon cycle trajectories and argue for more mechanistic models to improve predictive capabilities.The model used in the current study is available publicly (https://github.com/jinyun1tang/ECOSYS), and the current submission contains the python/ matlab codes for analyzing output from the model (includng a readme file to explain the codes). The benchmark data, also enclosed, was collected from a range of published and publicly available sources (extracted using GRABIT: https://www.mathworks.com/matlabcentral/fileexchange/7173-grabit). These sources describe warming induced changes in tundra/ boreal ecosystems.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS2015

An Assessment of Actual and Potential Building Climate Zone Change and Variability From the Last 30 Years Through 2100 Using NASA's MERRA and CMIP5 Simulations

Background: In the US, residential and commercial building infrastructure combined consumes about 40% of total energy usage and emits about 39% of total CO2 emission (DOE/EIA "Annual Energy Outlook 2013"). Building codes, as used by local and state enforcement entities are typically tied to the dominant climate within an enforcement jurisdiction classified according to various climate zones. These climate zones are based upon a 30-year average of local surface observations and are developed by DOE and ASHRAE. Establishing the current variability and potential changes to future building climate zones is very important for increasing the energy efficiency of buildings and reducing energy costs and emissions in the future. Objectives: This paper demonstrates the usefulness of using NASA's Modern Era Retrospective-analysis for Research and Applications (MERRA) atmospheric data assimilation to derive the DOE/ASHRAE building climate zone maps and then using MERRA to define the last 30 years of variability in climate zones for the Continental US. An atmospheric assimilation is a global atmospheric model optimized to satellite, atmospheric and surface in situ measurements. Using MERRA as a baseline, we then evaluate the latest Climate Model Inter-comparison Project (CMIP) climate model Version 5 runs to assess potential variability in future climate zones under various assumptions. Methods: We derive DOE/ASHRAE building climate zones using surface and temperature data products from MERRA. We assess these zones using the uncertainties derived by comparison to surface measurements. Using statistical tests, we evaluate variability of the climate zones in time and assess areas in the continental US for statistically significant trends by region. CMIP 5 produced a data base of over two dozen detailed climate model runs under various greenhouse gas forcing assumptions. We evaluate the variation in building climate zones for 3 different decades using an ensemble and quartile statistics to provide an assessment of potential building climate zone changes relative to the uncertainties demonstrated using MERRA. Findings and Conclusions: These results show that there is a statistically significant increase in the area covered by warmer climate zones and a tendency for a reduction of area in colder climate zones in some limited regions. The CMIP analysis shows that models vary from relatively little building climate zone change for the least sensitive and conservation assumptions to a warming of at most 3 zones for certain areas, particularly the north central US by the end of the 21st century.

Stackhouse, Paul W., Jr.↗

NASA NTRS2015

Statistical Analysis of Large Simulated Yield Datasets for Studying Climate Effects

Many studies have been carried out during the last decade to study the effect of climate change on crop yields and other key crop characteristics. In these studies, one or several crop models were used to simulate crop growth and development for different climate scenarios that correspond to different projections of atmospheric CO2 concentration, temperature, and rainfall changes (Semenov et al., 1996; Tubiello and Ewert, 2002; White et al., 2011). The Agricultural Model Intercomparison and Improvement Project (AgMIP; Rosenzweig et al., 2013) builds on these studies with the goal of using an ensemble of multiple crop models in order to assess effects of climate change scenarios for several crops in contrasting environments. These studies generate large datasets, including thousands of simulated crop yield data. They include series of yield values obtained by combining several crop models with different climate scenarios that are defined by several climatic variables (temperature, CO2, rainfall, etc.). Such datasets potentially provide useful information on the possible effects of different climate change scenarios on crop yields. However, it is sometimes difficult to analyze these datasets and to summarize them in a useful way due to their structural complexity; simulated yield data can differ among contrasting climate scenarios, sites, and crop models. Another issue is that it is not straightforward to extrapolate the results obtained for the scenarios to alternative climate change scenarios not initially included in the simulation protocols. Additional dynamic crop model simulations for new climate change scenarios are an option but this approach is costly, especially when a large number of crop models are used to generate the simulated data, as in AgMIP. Statistical models have been used to analyze responses of measured yield data to climate variables in past studies (Lobell et al., 2011), but the use of a statistical model to analyze yields simulated by complex process-based crop models is a rather new idea. We demonstrate herewith that statistical methods can play an important role in analyzing simulated yield data sets obtained from the ensembles of process-based crop models. Formal statistical analysis is helpful to estimate the effects of different climatic variables on yield, and to describe the between-model variability of these effects.

climate↗

NASA NTRS1997

Measurement of Geopotential Heights by GPS Radio Occultation

Geopotential heights of constant pressure surfaces are retrieved from global positioning system (GPS) radio occultation data. In order to assess accuracy a subset of data obtained by GPS/MET during spring 1995 and summer 1995 are compared to the output of the European Centre for Medium-Range Weather Forecasts (ECMWF) global model. The root-mean-square measurement error is 20 m throughout the upper troposphere and lower stratosphere. Furthermore, the ECMWF global model contains enhanced errors in the southeast Pacific. In probing the data for potential utility in climate studies, a Bayesian interpolation technique is used to map the geopotential height fields in the upper troposphere during the summer. Despite limitations of the GPS/MET data set the global average 300-mbar geopotential height over a 2-week period in summer 1995 is determined with an accuracy of 7 m. By obtaining greater coverage and partially resolving synoptic variability, a future constellation of 16 orbiting receivers could obtain global average geopotential height estimates in the upper troposphere with an accuracy of 1 m each day. Accuracy would be somewhat worse for regional studies, except in the tropics where synoptic variability is depressed.

Leroy, Stephen S.↗

NASA NTRS1985

Hemispheric differences in the earth radiation budget derived from Nimbus-7 related to climate studies

Earth Radiation Budget (ERB) data from Nimbus-7 over the period November 1978-June 1980 has consistently shown strong hemispheric differences when analyzed over different temporal and spatial scales. Hemispheric variations in time latitude cross sections of net and emitted radiation were found to be caused by changes in the earth-sun distance and continental effects. Maps of annual range for the entire earth calculated from monthly averages showed areas of high and low variability of the different ERB parameters. The ERB of these regional areas were examined and most of the variability was found to lay in the large amplitude of the annual solar cycle. Variations in the global annual cycle of albedo (Campbell et al. 1980) are studied with respect to differences in latitudinal averaged albedo. The anomaly in the annual cycle of global averaged albedo was found to be caused by tropical albedo changes.

Randel, D. L.↗

NASA NTRS1992

Tropical rainfall measuring mission - US/Japan science operations

The Tropical Rainfall Measuring Mission (TRMM) is a joint-venture between the National Aeronautics and Space Administration (NASA) and the National Space Development Agency (NASDA) of Japan. This cooperative program provides the resources required to place in a low-earth orbit a 3500 kilogram (kg) satellite. The satellite carries a complement of instruments designed to measure the tropical and subtropical rainfall as required to provide data products which will lead to climate modeling. This paper describes the mission, the spacecraft and instrument complement, and the data processing, distribution, and archiving system. Included in the paper is the possible, but not approved, an experimental data relay link using a NASDA data relay satellite.

Keating, Thomas↗

NASA NTRS2006

MSG SEVIRI Applications for Weather and Climate: Cloud Properties and Calibrations

SEVIRI data are cross-calibrated against the corresponding Aqua and Terra MODIS channels. Compared to Terra MODIS, no significant trends are evident in the 0.65, 0.86, and 1.6 micron channel gains between May 2004 and May 2006, indicating excellent stability in the solar-channel sensors. On average, the corresponding Terra reflectances are 12, 14, and 1% greater than the their SEVIRI counterparts. The Terra 3.8- micron channel brightness temperatures T are 7 and 4 K greater than their SEVIRI counterparts during day and night, respectively. The average differences between T for MODIS and SEVIRI 8.6, 10.8, 12.0, and 13.3- micron channels are between 0.5 and 2 K. The cloud properties are being derived hourly over Europe and, in initial comparisons, agree well with surface observations. Errors caused by residual calibration uncertainties, terminator conditions, and inaccurate temperature and humidity profiles are still problematic. Future versions will address those errors and the effects of multilayered clouds.

Minnis, Patrick↗

NASA NTRS2008

Climate Change Implications to Vegetation Production in Alaska

Investigation of long-term meteorological satellite data revealed statistically significant vegetation response to climate drivers of temperature, precipitation and solar radiation with exclusion of fire disturbance in Alaska. Abiotic trends were correlated to satellite remote sensing observations of normalized difference vegetation index to understand biophysical processes that could impact ecosystem carbon storage. Warming resulted in disparate trajectories for vegetation growth due to precipitation and photosynthetically active radiation variation. Interior spruce forest low lands in late summer through winter had precipitation deficit which resulted in extensive fire disturbance and browning of undisturbed vegetation with reduced post-fire recovery while Northern slope moist alpine tundra had increased production due to warmer-wetter conditions during the late 1990s and early 2000s. Coupled investigation of Alaska s vegetation response to warming climate found spatially dynamic abiotic processes with vegetation browning not a result from increased fire disturbance.

Neigh, Christopher S.R.↗

DOE OSTI2025

Degradation and deforestation increase the sensitivity of the Amazon Forest to climate extremes

About 40% of the Brazilian Amazon has been deforested or suffered changes in forest structure through degradation (selective logging, fires, and fragmentation). The impact of forest degradation on the forest’s sensitivity to climate extremes has not been fully explored because of a lack of data and the complex interplay of forest structure and climate drivers. Here, we combined forest structure data from 545 airborne lidar transects (375 ha each) across the Brazilian Amazon with the Ecosystem Demography Model (ED2). We explore the forest’s functional response to near-present (1981–2019) climate extremes under observed forest structure from lidar ( Control ) and two forest structure change scenarios: (1) forest recovery by excluding all future deforestation and degradation ( Recovery ) and (2) expansion of selective logging and deforestation ( Degradation ). Using the Control simulation, we found a close and positive association between local forest aboveground biomass and the predicted gross primary productivity (GPP) and evapotranspiration (ET). Moreover, both GPP and ET respond negatively to extremes in vapor pressure deficit and downwelling shortwave irradiance in degraded forests in Eastern and Southern Amazon, indicating high sensitivity to droughts. Locally high-biomass forest patches showed little or no negative response of GPP and ET to extreme drought conditions whereas low-biomass forest patches in the same locations—typically degraded forest canopies—responded negatively to higher moisture stress. The results from the Recovery scenario showed similar results to simulations with observed structure; however, under the Degradation scenario, low-biomass forest patches became more abundant, resulting in more regions where GPP and ET are negatively impacted by hot drought conditions according to the ED2 model. Our results suggest that local forest structure is a critical determinant of an ecosystem’s response to climate variability, and that the loss of canopy trees in the Amazon through forest degradation could increase and expand forest vulnerability to droughts.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS2006

The NPOESS Preparatory Project Science Data Segment: Brief Overview

The NPOESS Preparatory Project (NPP) provides remotely-sensed land, ocean, atmospheric, ozone, and sounder data that will serve the meteorological and global climate change scientific communities while also providing risk reduction for the National Polar-orbiting Operational Environmental Satellite System (NPOESS), the U.S. Government s future low-Earth orbiting satellite system monitoring global weather and environmental conditions. NPOESS and NPP are a new era, not only because the sensors will provide unprecedented quality and volume of data but also because it is a joint mission of three federal agencies, NASA, NOAA, and DoD. NASA's primary science role in NPP is to independently assess the quality of the NPP science and environmental data records. Such assessment is critical for making NPOESS products the best that they can be for operational use and ultimately for climate studies. The Science Data Segment (SDS) supports science assessment by assuring the timely provision of NPP data to NASA s science teams organized by climate measurement themes. The SDS breaks down into nine major elements, an input element that receives data from the operational agencies and acts as a buffer, a calibration analysis element, five elements devoted to measurement based quality assessment, an element used to test algorithmic improvements, and an element that provides overall science direction. This paper will describe how the NPP SDS will leverage on NASA experience to provide a mission-reliable research capability for science assessment of NPP derived measurements.

Schweiss, Robert J.↗

NASA NTRS2008

Space-based Ornithology-Studying Bird Migration and Environmental Change in North America

Natural fluctuations in the availability of critical stopover sites coupled with anthropogenic destruction of wetlands, land-use change, and anticipated losses due to climate change present migratory birds with a formidable challenge. We have developed an individual-based, spatially explicit bird migration model that simulates the migration routes, timing and energy budgets of individual birds under dynamic weather and land surface conditions. Our model incorporates biophysical constraints, individual bird energy status, bird behavior, and flight aerodynamics. We model the speed, direction, and timing of individual birds moving through a user specified Lagrangian grid. The model incorporates environmental properties including wind speed and direction, topography, dynamic hydrologic properties of the landscape, and environmental suitability. The model is driven by important variables estimated from satellite observations of the land surface, by data assimilation products from weather and climate models, and biological field data. We illustrate the use of the model to study the impact of both short- and long-term environmental variatios, e.g. climate, drought, anthropogenic, on migration timing (phenology), spatial pattern, and fitness (survival and reproductive success). We present several theoretical simulations of the spring migration of Pectoral Sandpiper (Calidris melanotos) in North America with emphasis on the Central flyway from the Gulf of Mexico to Alaska.

Smith, James↗

NASA NTRS2016

A Model-Model and Data-Model Comparison for the Early Eocene Hydrological Cycle

A range of proxy observations have recently provided constraints on how Earth's hydrological cycle responded to early Eocene climatic changes. However, comparisons of proxy data to general circulation model (GCM) simulated hydrology are limited and inter-model variability remains poorly characterised. In this work, we undertake an intercomparison of GCM-derived precipitation and P - E distributions within the extended EoMIP ensemble (Eocene Modelling Intercomparison Project; Lunt et al., 2012), which includes previously published early Eocene simulations performed using five GCMs differing in boundary conditions, model structure, and precipitation-relevant parameterisation schemes. We show that an intensified hydrological cycle, manifested in enhanced global precipitation and evaporation rates, is simulated for all Eocene simulations relative to the preindustrial conditions. This is primarily due to elevated atmospheric paleo-CO2, resulting in elevated temperatures, although the effects of differences in paleogeography and ice sheets are also important in some models. For a given CO2 level, globally averaged precipitation rates vary widely between models, largely arising from different simulated surface air temperatures. Models with a similar global sensitivity of precipitation rate to temperature (dP=dT ) display different regional precipitation responses for a given temperature change. Regions that are particularly sensitive to model choice include the South Pacific, tropical Africa, and the Peri-Tethys, which may represent targets for future proxy acquisition. A comparison of early and middle Eocene leaf-fossil-derived precipitation estimates with the GCM output illustrates that GCMs generally underestimate precipitation rates at high latitudes, although a possible seasonal bias of the proxies cannot be excluded. Models which warm these regions, either via elevated CO2 or by varying poorly constrained model parameter values, are most successful in simulating a match with geologic data. Further data from low-latitude regions and better constraints on early Eocene CO2 are now required to discriminate between these model simulations given the large error bars on paleoprecipitation estimates. Given the clear differences between simulated precipitation distributions within the ensemble, our results suggest that paleohydrological data offer an independent means by which to evaluate model skill for warm climates.

Boundary conditions↗

NASA NTRS2006

Application of the Convolution Formalism to the Ocean Tide Potential: Results from the Gravity and Recovery and Climate Experiment (GRACE)

A computationally efficient approach to reducing omission errors in ocean tide potential models is derived and evaluated using data from the Gravity Recovery and Climate Experiment (GRACE) mission. Ocean tide height models are usually explicitly available at a few frequencies, and a smooth unit response is assumed to infer the response across the tidal spectrum. The convolution formalism of Munk and Cartwright (1966) models this response function with a Fourier series. This allows the total ocean tide height, and therefore the total ocean tide potential, to be modeled as a weighted sum of past, present, and future values of the tide-generating potential. Previous applications of the convolution formalism have usually been limited to tide height models, but we extend it to ocean tide potential models. We use luni-solar ephemerides to derive the required tide-generating potential so that the complete spectrum of the ocean tide potential is efficiently represented. In contrast, the traditionally adopted harmonic model of the ocean tide potential requires the explicit sum of the contributions from individual tidal frequencies. It is therefore subject to omission errors from neglected frequencies and is computationally more intensive. Intersatellite range rate data from the GRACE mission are used to compare convolution and harmonic models of the ocean tide potential. The monthly range rate residual variance is smaller by 4-5%, and the daily residual variance is smaller by as much as 15% when using the convolution model than when using a harmonic model that is defined by twice the number of parameters.

models↗

NASA NTRS2019

North American Extreme Precipitation Events and Related Large-Scale Meteorological Patterns: a Review of Statistical Methods, Dynamics, Modeling, and Trends

This paper surveys the current state of knowledge regarding Large-Scale Meteorological Patterns (LSMPs) associated with short-duration (less than one week) extreme precipitation events over North America. In contrast to teleconnections, which are typically defined based on the characteristic spatial variations of a meteorological field or on the remote circulation response to a known forcing, LSMPs are defined relative to the occurrence of a specific phenomenon—here, extreme precipitation—and with an emphasis on the synoptic scales that have a primary influence in individual events, have medium-range weather predictability, and are well-resolved in both weather and climate models. For the LSMP relationship with extreme precipitation, we consider the previous literature with respect to definitions and data, dynamical mechanisms, model representation, and climate change trends. There is considerable uncertainty in identifying extremes based on existing observational precipitation data and some limitations in analyzing the associated LSMPs in reanalysis data. Many different definitions of “extreme” are in use, making it difficult to directly compare different studies. Dynamically, several types of meteorological systems—extratropical cyclones, tropical cyclones, mesoscale convective systems, and mesohighs—and several mechanisms—fronts, atmospheric rivers, and orographic ascent—have been shown to be important aspects of extreme precipitation LSMPs. The extreme precipitation is often realized through mesoscale processes organized, enhanced, or triggered by the LSMP. Understanding of model representation, trends, and projections for LSMPs is at an early stage, although some 4 promising analysis techniques have been identified and the LSMP perspective is useful for evaluating model dynamics.

Mathew Barlow↗

NASA NTRSDate not supplied

Web-based Visualization and Analytics of Petascale Data: Equity as a Tide that Lifts All Boats

Scientists generate petabytes of data daily to help uncover environmental trends or behaviors that are hard to predict. For example, understanding climate simulations based on the long-term average of temperature, precipitation, and other environmental variables is essential to predicting and establishing root causes of future undesirable scenarios and assessing possible mitigation strategies. While supercomputer centers provide a powerful infrastructure for generating petabytes of simulation output, accessing and analyzing these datasets interactively remains challenging on multiple fronts. This paper presents an approach to managing, visualizing, and analyzing petabytes of data within a browser on equipment ranging from the top NASA supercomputer to commodity hardware like a laptop. Our novel data fabric abstraction layer allows user-friendly querying of scientific information while hid-ing the complexities of dealing with file systems or cloud services.We also optimize network utilization while streaming from petas-cale repositories through state-of-the-art progressive compression algorithms. Based on this abstraction, we provide customizable dashboards that can be accessed from any device with any inter-net connection, enabling interactive visual analysis of vast amounts of data to a wide range of users - from top scientists with access to leadership-class computing environments to undergraduate students of disadvantaged backgrounds from minority-serving institutions. We focus on NASA’s use of petascale climate datasets as an example of particular societal impact and, therefore, a case where achieving equity in science participation is critical. We validate our approach by improving the ability of climate scientists to visually explore their data via two fully interactive dashboards. We further validate our approach by deploying the dashboards and simplified training materials in the classroom at a minority-serving institution.These dashboards, released in simplified form to the general public, contribute significantly to a broader push to democratize the access and use of climate data.

Data visualization↗

NASA NTRS2021

Global Soil Water Estimates as Landslide Predictor: The Effectiveness of SMOS, SMAP, and GRACE Observations, Land Surface Simulations, and Data Assimilation

This global feasibility study assesses the potential of coarse-scale, gridded soil water estimates for the probabilistic modeling of hydrologically-triggered landslides, using Soil Moisture Ocean Salinity (SMOS), Soil Moisture Active Passive (SMAP) and Gravity Recovery and Climate Experiment (GRACE) remote sensing data, Catchment Land Surface Model (CLSM) simulations and six data products based on the assimilation of SMOS, SMAP, and/or GRACE observations into CLSM. SMOS or SMAP observations (~40-km resolution) are only available for less than 20% of the globally reported landslide events, because they are intermittent and uncertain in regions with complex terrain. GRACE terrestrial water storage estimates include 75% of the reported landslides but have coarse spatial and temporal resolutions (monthly, ~300-km). CLSM soil water simulations have the added advantage of complete spatial and temporal coverage, and are found to be able to distinguish between “stable slope” (no landslide) conditions and landslide-inducing conditions in a probabilistic way. Assimilating SMOS and/or GRACE data increases the landslide probability estimates based on soil water percentiles for the reported landslides, relative to model-only estimates at 36-km resolution for the period 2011-2016, unless the CLSM model-only soil water content is already high (≥ 50th percentile). The SMAP Level 4 data assimilation product (at 9-km resolution, period 2015-2019) more generally updates the soil water conditions towards higher landslide probabilities for the reported landslides, but is similar to model-only estimates for the majority of landslides where SMAP data cannot easily be converted to soil moisture owing to complex terrain.

Anne Felsberg↗

NASA NTRS2016

IN11B-1621: Quantifying How Climate Affects Vegetation in the Amazon Rainforest

Amazon droughts in 2005 and 2010 have raised serious concern about the future of the rainforest. Amazon forests are crucial because of their role as the largest carbon sink in the world which would effect the global warming phenomena with decreased photosynthesis activity. Especially, after a decline in plant growth in 1.68 million km2 forest area during the once-in-a-century severe drought in 2010, it is of primary importance to understand the relationship between different climatic variables and vegetation. In an earlier study, we have shown that non-linear models are better at capturing the relation dynamics of vegetation and climate variables such as temperature and precipitation, compared to linear models. In this research, we learn precise models between vegetation and climatic variables (temperature, precipitation) for normal conditions in the Amazon region using genetic programming based symbolic regression. This is done by removing high elevation and drought affected areas and also considering the slope of the region as one of the important factors while building the model. The model learned reveals new and interesting ways historical and current climate variables affect the vegetation at any location. MAIAC data has been used as a vegetation surrogate in our study. For temperature and precipitation, we have used TRMM and MODIS Land Surface Temperature data sets while learning the non-linear regression model. However, to generalize the model to make it independent of the data source, we perform transfer learning where we regress a regularized least squares to learn the parameters of the non-linear model using other data sources such as the precipitation and temperature from the Climatic Research Center (CRU). This new model is very similar in structure and performance compared to the original learned model and verifies the same claims about the nature of dependency between these climate variables and the vegetation in the Amazon region. As a result of this study, we are able to learn, for the very first time how exactly different climate factors influence vegetation at any location in the Amazon rainforests, independent of the specific sources from which the data has been obtained.

global warming↗