---
title: Exogenous-Only Temperature Forecasting
url: https://www.emergentmind.com/topics/exogenous-only-temperature-forecasting
type: topic
---

# Exogenous-Only Temperature Forecasting

Exogenous-only temperature forecasting refers to quantitative prediction of future air or surface temperatures using only external (exogenous) variables—excluding any direct autoregressive (past value) information from the target temperature series. All relevant exogenous-only approaches strictly prohibit the algorithm from “seeing” or conditioning on past temperatures of the location being forecast, either during training or inference. This field has gained increasing attention due to operational constraints (e.g., absence of reliable on-site past temperature data, need for robustness in nonstationary or regime-shifting settings), the desire to avoid artificial self-persistence, and the opportunity to exploit teleconnections and high-fidelity weather and climate model outputs.

## 1. Fundamental Principles of Exogenous-Only Temperature Forecasting

The defining characteristic of exogenous-only forecasting is the strict partition of model inputs into:
- Exogenous variables: meteorological or environmental covariates (e.g., ensemble NWP or reanalysis fields, observed weather variables at other locations, climate indices, etc.).
- Target: the temperature series to be forecast, which is inaccessible at inference except for historical “ground truth” labels.

Exogenous-only approaches are pursued either for scientific purposes (causal analysis, teleconnection modeling) or necessitated by data availability. By construction, they avoid recursiveness: there are no lagged observations of the predictand in the predictor set. This rules out any ARIMA or classic autoregressive DL sequence models in their vanilla form. The result is models that can only exploit the information content present in the selected exogenous variables.

Non-parametric analog methods, linear regression, regularized ensemble postprocessors, convolutional neural networks, and transformer-free deep architectures have been investigated, all subject to this constraint [1507.03283, 2211.15856, 2205.10972, 2406.02141, 2505.23116, 2512.10866].

## 2. Exogenous Predictors and Feature Engineering

Exogenous-only frameworks rely on an expanded set of external covariates:
- Synoptic and climate-scale predictors: surface and upper-air fields from NWP/reanalysis (T2m, Z500, SST, soil moisture), indices such as ENSO, SOI, multivariate MEI, Darwin/Tahiti sea level pressure, etc. [1507.03283, 2211.15856, 2205.10972].
- NWP or weather model outputs: deterministic or ensemble model forecasts for temperature (at any grid, not the target location), dew point, humidity, wind, precipitation, geopotential, etc. [2211.15856, 2406.02141].
- Local environmental exogenous inputs: observations at remote sites and physical covariates (e.g., relative humidity, barometric pressure, radiation, precipitation, cloud cover, wind), but not the history of the target temperature [2505.23116, 2512.10866].
- Positional and static features: latitude, longitude, elevation, station/category embeddings to encode location-specific priors; time-of-day, day-of-year, lead time conditioning [2406.02141].

Principal component or patch-based representations are used for high-dimensional global predictors such as SST [2211.15856]. Feature preprocessing includes z-score standardization or instance normalization per channel, log-transformation of certain physical variables, and bias correction for NWP gridpoint forecasts [2406.02141, 2505.23116, 2512.10866].

## 3. Model Architectures and Algorithmic Strategies

A wide spectrum of architectures has been deployed, reflecting the absence of autoregressive target information.

### a. Linear and Generalized Linear Models
- Pure one-shot linear regression on windowed exogenous histories [2512.10866].
- Bias-corrected, trend/seasonal-decomposed linear projections (NLinear, DLinear) to mitigate nonstationarity and capture additive structure [2512.10866].
- Linear regression per-location or as a global estimator, sometimes with ensemble-member expansion [2211.15856, 2406.02141].

### b. Nonparametric Analog and Rule-Based Methods
- Predictor-sum statistics for pairs of exogenous variables (analog approach): compute weighted sums of standardized anomalies at remote sites over a lead time and summation window, select those that historically maximized anticipation of target extremes [1507.03283].
- Teleconnection mining via a data-driven analog-configuration search, bypassing classical parametric fit to extremes [1507.03283].

### c. Deep Learning Architectures
- CNN/U-Net architectures for spatial map-to-map regression from exogenous field tensors at initialization to temperature anomalies at horizon, trained via MSE or super-quadratic losses targeting extreme temperature events [2205.10972].
- Multi-layer perceptrons (MLP) with station and lead-time embeddings, trained for parameter prediction (e.g., to output location/time-dependent parameters of probabilistic distributions) [2406.02141].
- Cross-correlation embedding models that couple exogenous features in a time-invariant, linear fashion using shared 1D convolutions, followed by patch embedding and global linear heads for both short-term and long-term exogenous-only forecasting. Residual endogenous channels can be removed entirely in strict exogenous-only regimes [2505.23116].

### d. Ensemble and Probabilistic Postprocessing
- Quantile regression forests, Bernstein polynomial quantile networks, normal-parametric NNs for full-distribution prediction from raw deterministic or ensemble NWP outputs, with or without explicit model spread information [2406.02141, 2211.15856].
- Stacked models blending predictions from linear, RF, and CNN architectures as base learners and meta-learners for ensembling [2211.15856].

### e. Transformer-type vs. Linear Models
- Standard and enhanced Transformer architectures (Informer, Autoformer) compared to linear baselines for exogenous-only indoor temperature forecasting, with empirical findings that linear architectures (NLinear, DLinear) are substantially more robust and accurate in these scenarios [2512.10866].

## 4. Training, Validation, and Loss Functions

Exogenous-only models are trained exclusively on historical exogenous/target pairs, never feeding back predictions. Key regimes:
- Strict temporal splits, blocking all future-leaking variables and avoiding cross-talk with target temperature sequences [1507.03283, 2512.10866].
- Windowing: Input look-back of length L in exogenous features, forecast horizon H, building non-overlapping [L→H] training blocks [2512.10866].
- Bias correction, outlier removal (>15 K), z-score or min-max scaling restricted to training data for robust normalization [2406.02141, 2505.23116].

Losses are selected based on the output endpoint:
- L2 (MSE) for mean prediction.
- Symmetric exponential or "pinball" (quantile) loss for extreme-value targeting [2205.10972, 2211.15856].
- Continuous Ranked Probability Score (CRPS) for probabilistic output calibration [2406.02141].
For analog methods, combinatorial search is guided by rule quality criteria based on frequency and sharpness of predictor-exceedance during true extremes [1507.03283].

Early stopping, batch normalization, regularization (e.g., weight decay), and ensembling over model seeds are applied as needed. Station and lead-time embeddings are learned via additive parametric vectors [2406.02141].

## 5. Performance Benchmarks and Empirical Insights

Across operational, climate, and energy/building datasets, exogenous-only methods exhibit several robust findings.

| Approach           | Domain/Horizon                | Key Metrics / Results                                | Reference        |
|--------------------|------------------------------|------------------------------------------------------|------------------|
| Nonanticipative Analog | Daily/annual, US, extremes    | Hit rate for hot-waves up to 18.2%, 100% sign accuracy | [1507.03283]     |
| Neural Weather Models | Global, 1-28 days             | ACC at 14d=0.68 (vs persist 0.30, ECMWF 0.62); HR≈0.55 | [2205.10972]     |
| Deterministic-to-Prob. NN | In-situ, 10-day, N. America | CRPS reduced by 15.6% vs. naive; best with BQN/DRN    | [2406.02141]     |
| CrossLinear (pure exo)   | Real-world, var. term        | Consistent improvements over neural/attention models   | [2505.23116]     |
| LTSF-Linear/Transformer  | Indoor, 4-day                | NLinear MAE=0.2461, DLinear=0.2811 outperforms Transformer (MAE>0.8) | [2512.10866] |
| ML postprocessed ensembles | CONUS, subseasonal          | Stacked (full ensemble) R² ≈ 0.18, RMSE=1.77°C        | [2211.15856]     |

A consistent pattern emerges: well-designed linear and convolutional models often outperform deeper attention-based architectures under pure exogenous-only restrictions, due to their regularization and tendency to avoid overparameterization and overfitting [2512.10866]. Use of full ensemble vectors as exogenous features systematically outperforms summarization (e.g., ensemble mean), with statistically significant skill improvements [2211.15856].

For extremes, analog and exponential-loss-trained CNNs yield meaningful hit rates, but challenges in recall and smoothing remain, especially at long leads [1507.03283, 2205.10972]. Probabilistic postprocessing of deterministic forecasts via NNs, MLPs, or ensemble methods can recover sharp, calibrated predictive distributions without costly ensemble simulation [2406.02141].

## 6. Operational Implications, Strengths, and Limitations

Exogenous-only approaches exhibit several distinctive operational and scientific properties:
- No artificial self-persistence: Models avoid overfitting to slow-moving internal dynamics or persistent anomalies in the target, as only exogenous drivers are used [1507.03283].
- Robustness to out-of-distribution and nonstationary changes, assuming exogenous predictors contain up-to-date causal or correlational information [2512.10866].
- Low computational cost for probabilistic prediction: Single deterministic NWP or ML-based forecasts can be efficiently postprocessed into full probabilistic distributions [2406.02141].
- Teleconnection and remote driver identification, particularly in analog and rule-mining methods, which automatically discover historical precedents without forced parametric form [1507.03283].
- Limitations include reduced “upper bound” on forecast skill (modest hit rates for extremes, e.g. 10–20%), susceptibility to loss in long-term state memory (soil moisture feedback, persistent local anomalies), and, for DL models, the risk of over-smoothing or amplitude underestimation at long leads [2205.10972, 2406.02141].

Empirical findings underscore the perils of high-capacity models (e.g., Transformer-family) in overfitting or collapsing on purely exogenous input, whereas explicit linear decomposition or convolutional inductive biases yield more consistent performance [2505.23116, 2512.10866].

## 7. Research Developments and Directions

Recent developments have advanced exogenous-only temperature forecasting in several respects:
- Plug-and-play cross-correlation embedding approaches for modular integration of exogenous-only forecasting into broader time series pipelines, with explicit removal of shallow autoregressive dynamics [2505.23116].
- Evaluation frameworks and community benchmarks (“UrbanAI 2025 Challenge”) enforcing strict exogenous-only constraints for energy/building and indoor-outdoor forecasting, revealing the practical supremacy of normalized and decomposed linear models [2512.10866].
- Comprehensive postprocessing frameworks capable of upgrading deterministic NWP and ML forecasts to well-calibrated probabilistic outputs—spanning static (EMOS), quantile-based (BQN), and neural density methods [2406.02141].
- Exploitation of the full ensemble-member structure in subseasonal and seasonal climate prediction tasks, yielding skill advances attributable to high-dimensional exogenous signals and auxiliary variables [2211.15856].
- Persistent open problems include scaling hit rates for extreme events, reducing computational search burden (in analog/rule-based systems), adaptively expanding exogenous pools (via feature selection, clustering, or representation learning), and understanding limits of predictability in the absence of endogenous data.

Together, the exogenous-only paradigm, as developed and validated in these studies, provides a rigorous discipline for the design, evaluation, and interpretation of temperature forecasts—a framework that, while sometimes sacrificing maximum recall, offers critical benefits in transparency, transferability, and real-world operational reliability.

Source: https://www.emergentmind.com/topics/exogenous-only-temperature-forecasting